London Borough of Southwark v IBM UK Ltd [2011] EWHC 549 (TCC)

[2011] EWHC 549 (TCC)Case No HT-09-325
IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
TECHNOLOGY AND CONSTRUCTION COURT
Venue Royal Courts of Justice, Strand, London, WC2A 2LLDate 17 th March 2011
Before
MR JUSTICE AKENHEAD
Between
THE MAYOR AND BURGESSES OF THE LONDON BOROUGH OF SOUTHWARKClaimantIBM UK LIMITEDDefendant
Nicholas Stewart QC and Simon Perhar (instructed by the Legal Services Department of the London Borough of Southwark) for ClaimantJeremy Nicholson QC and Terence Bergin (instructed by Blake Lapthorn) for DefendantHearing Hearing dates: 7-10, 14, 15 and 17 February 2011
JUDGMENT

Introduction

[1]These proceedings relate to the provision of software and associated consultancy services by IBM UK Ltd ("IBM") to the London Borough of Southwark ("Southwark") in 2006 and 2007. The software was to be provided in connection with the implementation of a Master Data Management ("MDM") system within Southwark which would rationalise a large amount of information historically and continuously logged on to a number of different computer systems within Southwark.[2]Southwark complains that the software provided and in particular what was called Arcindex, manufactured and marketed by a different company, Orchard, was unsatisfactory and otherwise unfit for purpose. Southwark’s claim was put in a variety of ways ranging from misrepresentation, breach of a collateral warranty, negligence and breach of one of the three specific contracts entered into between the parties, although the first three of these grounds were expressly abandoned by Counsel for Southwark in closing. Having been provided in late 2006 and up to about June 2007, the software and services provided by IBM were abandoned by Southwark in the second half of 2007. The Background

Sub-issues are:

[3]As a highly populated London borough, Southwark has and has had a large number of properties and people within the borough and contact between the borough and the people who live and work in it is through a number of points of contact. For instance, Council Tax payers and business rates payers have to be identified and billed. It is necessary to keep extensive records and most of these are kept on a number of disparate computerised systems, disparate in the sense that each of them deals with different subject matters. These included systems (sometimes called “source systems”) for Customer Relationship Management ("CRM”), Strategic Services Application (“eReg”), National Non-Domestic Rates ("NNDR”), Local Land and Property Gazetteer (“LLPG”), Housing Maintenance Application (“iWorld”) and Council Tax (“CTAX”). Unsurprisingly, over the years a mass of information has been computerised about people, property and businesses within the borough and, equally unsurprisingly, some people, businesses or property are entered by mistake more than once within the same source system; sometimes, there are discrepancies between the names, properties and other details in the different entries. Sometimes, records on one system record different information about the same person and property. These mistakes arise from many reasons, which include incorrect information provided or transposed.[4]The concept of MDM by way of a central computerised system has gathered in popularity and for some 10 years computer companies, such as IBM, have invested research and development of, and cost in acquiring, software and hardware to accommodate the aspirations of potential clients in this context. The main objective of MDM is to try to harmonise and improve data. As the concept has developed, the heart of MDM is a central index which in broad terms signposts all of the data within an organisation located within the different source systems; that central index provides, houses or procures a master data set which can be accessed to obtain consistent and accurate master data. At least in so far as MDM related to management of local authority data, MDM systems were in 2006, if not in their infancy in terms of development, at least in the relatively early stages of development. The market for MDM was described as being immature in 2006. Indeed, until IBM acquired the “Ascential” products from another organisation, it did not have a viable MDM capacity. In practice, it has been found by a number of the London boroughs which have introduced or tried to introduce MDM systems that they are complex. In the Best Practice Guide for MDM Implementations issued by DataConnects published in June 2008, the following was said: “Time & Resources Be under no illusion, projects of this nature take time, do not underestimate the number of resources and time required, estimate how long you think it will take and triple it and then double that. The perception by many is the matching of data and resources required should take one or two person[s] just a few weeks to sort out, this simply isn't true. Projects of this nature are complex, often due to the data quality issues they revealed, they often serve to highlight how poor current processes and practices are. It can be often likened to opening a can of worms that cannot be ignored once opened, revealing disasters waiting to happen. Costs Costs Some LAs reported that additional costs were incurred due to:• . The immaturity of the matching product purchase, extra money, time and resource was required to develop software to cope with product deficiencies and test product enhancements• . Investment in operator skill set was far higher than anticipated• . A higher number of matches than anticipated that required manual resolution."[5]Southwark can be classified as one of the major London boroughs, having in 2006 to 2007 expenditure of more than £1.2 billion. It had a separate Information Technology Department, staffed by experienced IT managers and other professionals, some directly employed and others engaged as consultants. It outsourced a number of its IT operations to organisations such as Pearson (later Vangent) and Serco.[6]In about 2005, Southwark was considering the introduction of an MDM system. At this stage, a number of key personnel were involved in this project. David Currey, initially a programme manager for systems at Southwark, became the head of information services in December 2005. Mike Katz was an independent consultant who had been retained as a project manager since 2001. James Orrom was probably also an independent consultant with experience in IT matters. Initially, Southwark approached SAP (UK) Ltd which was part of a German group and Southwark with SAP which drew up a document entitled "SAP Master Data Management Phase 1 Project Brief” (“The SAP Brief”), the version in the trial papers being dated 13 March 2006. This was a high level and non-detailed document and is written in a language in part at least familiar to those involved in this type of project:
“1 Overview Master Data Management refers to key reference information about customers, products, vendors, employees, organisational units, accounts, etc. Master Data Management is about bringing the multiple variances of the key data elements into a single harmonised data set so that everyone has the same view of the world. Once harmonised this key data is needed by an organisation at an operational level and from a leadership perspective. It allows consistent answers to management questions… The council had recently implemented a Customer Service Centre in partnership with Pearson eGovernment Solution which is supported by a SAP Customer Relationship Management (CRM) System. The CRM system holds data on the Council’s customers and properties. Each back office department within the Council also holds the same customer and property data multiple times in differing formats and differing states of completeness/accuracy in the many IT systems that support the delivery of services across the organisation. The Problem: • Data on a single customer/property is held multiple times in the same system. • Data on a single customer/property is maintained in multiple systems in multiple formats. • Data on a single customer/property is going to be held in varying degrees of accuracy and consistency. • The Council needs to understand the relationship that a customer has with the property in order to improve service delivery. • It is not possible to deliver truly joined up service delivery if each of the departments/services are using different instances of the Master Customer. The Solution: • Bring the multiple customer/property data sets together, using MDM cleanse and match each data into two distinct Master Data sets the customers, property and establish a relationship/link between them…. 2.1 Project Objectives The purpose of this document is to define the scope Phase 1 of this project so that a total fixed price can be calculated enabling LBS to make a checkpoint Go/go decision for this project. The project will decide whether to proceed on the merits on the benefit is realised in Phase 1 alone, therefore the deliverables in Phase 1 must be achievable and provide sufficient business benefit. Project Objectives: To obtain a complete master data set for customers and properties within Southwark. To provide a system for maintaining (and where required establishing) the relationships between customers and properties. To realise operational efficiencies in the maintenance of Master Data. To provide the infrastructure for ID Management/Authentication for customer self-service through Government Connect. To provide the capability for a 360 degree view of the customer within the Data Warehouse. 3. Project Scope 3.1 …It it should be understood that the scope of this project is to build the basis for developing a consistent master data set for the entire organisation. It should not be assumed that this initial project would encompass all the business process changes required. This project will not be implemented in a big bang approach but rather through an iterative process solving a specific business need… Outline Phase 1 Deliverables Must Have: 1) Infrastructure & environment setup of MDM components in the current LDS landscape 2) Configuration/ Development of the property data model/scheme in MDM for Southwark properties to define… 3) Configuration/Development of the customer data model/scheme in MDM for Customers… 4) Data consolidation identification & merging definition… 5) Data Cleansing & consolidation of data imported from LBS Strategic applications: 6) Data rules, validation duplication definition 7) People property linkage & integration in the MDM Data model 8) Import customer data from the following 4 strategic applications for Southwark Master customer data set: eReg… CRM… NNDR… iWorld… 9) Import property data from the following 4 strategic applications for Southwark Master property data set: LLPG… CRM… iWorld… 16) Definition of MDM user roles & authorisations… Would Like To Have: 18) Using the cleansing and de-duplicating functionality of MDM merge where possible all duplicate customer records and reassign service tickets…”
[7]There was a workshop held as between SAP and the Southwark MDM project team on 29 March 2006 at which, the Notes thereof reveal, there was a wide ranging discussion about what was required. The Notes reveal that there was some discussion about "dealing with duplicates", fixing them back at the source systems and then feeding them back into the MDM system. The problem perceived by Southwark (which was actually present) was the probable large quantity of duplicate records both within individual source systems and as between different source systems. Thus, a Mr A Smith of 23A, Acacia Avenue SE1 2AS could appear on two records, one in that form and another as Anthony Smith of the same address. Another problem is that for one reason or another either or both might contain errors so that for instance the second record might identify Mr Smith as Mr A Smythe or the postcodes could be different. There would be some instances when the same person or property appeared on three or more records created on his or her behalf. Apart from cluttering up the systems, there would be room for confusion with some information stored on one record but not on another for the same person.[8]By March 2006, IBM was marketing under a new name a software package which had previously been marketed by Ascential which it had taken over some months before. IBM called it “Websphere” and the three key software elements were: (a)”ProfileStage” which had the capability to look at data stored across a range of source systems, to look at data in any form and to bring groups of data together. (b) “DataStage” which had capability to load data into a separate system or a core system. (c) “QualityStage” which had the capability to apply structure to data by categorising and profiling it. 12. The Evidence[9]I heard evidence from one factual witness from Southwark, Clare Troy, who was retained by Southwark as a consultant data integrity specialist on 22 June 2007 and who was intimately involved with considering over the following few months what had been provided by IBM; she was ultimately instrumental in bringing about Southwark’s decision to abandon work and software provided by IBM. Perhaps surprisingly in the light of the claims for misrepresentation and collateral warranty, Southwark called no factual witness who was involved in 2006 or otherwise in the procurement of the IBM software and services. There has been no suggestion that Messrs Katz, Orrom, Currey or a Ms Leahy who was in charge of the implementation of the procurement for Southwark are not available and could not have been called. Whilst I will draw no negative inference as a result of these witnesses not being called, Southwark does not have the benefit of their evidence and that may count against them in the sense that the burden of proof is primarily upon it.[10]IBM called five witnesses of fact, four from IBM and one from Orchard. Mr Thomson was previously employed in a sales capacity by Ascential, acquired by IBM in November 2005; Mr Thomson then joined IBM in a similar capacity and was involved in the initial contacts between Southwark and IBM. Mr Burgess of IBM was a Specialist Sales Representative who was involved in the sale of the IBM products to Southwark primarily in 2006 although he had some involvement in 2007. Mr Keith was the "on Demand Architect” who was also involved in 2006 at meetings between IBM and Southwark. Mr Avon was an IBM services manager who was also involved in the sales meetings with Southwark but was also involved in the installation of the software in 2007 and the addressing of complaints from Southwark.[11]Each side called a computer expert, Southwark calling Annette Hughes and IBM Jason Coyne. There has been no suggestion that they were anything other than impartial and qualified to give evidence. Both have had to some extent to attempt to reconstruct what happened and what Southwark’s requirements and were or may have been. Ms Hughes has not worked on Arcindex whilst Mr Coyne has, which in my view gives him some greater reliability as to what Arcindex can and cannot do.[12]I found all the factual witnesses to be honest and helpful. So far as the experts are concerned, I have to assess their evidence by reference to the facts which I find in relation to what Southwark’s requirements really were. The Involvement of IBM[13]Mr Katz and Mr Orrom, who were actually and apparently well qualified and experienced in matters relating to computers, software and systems, began to have doubts about whether the proposed SAP systems were appropriate albeit that Mr Currey was confident about them. It has not been established who contacted whom but contact was made between Southwark and, initially, a Ms Jones of IBM in April 2006. On about 12 April 2006, she was sent a copy of the SAP Brief which she passed to Mr Thomson and Mr Burgess to look at, which they did. Southwark was not a previous customer of IBM. An internal conference was set up within IBM organised by Ms Jones and attended by Messrs Thomson, Burgess and Keith. They reviewed briefly what the client’s requirements were and it occurred to them that the Websphere products newly marketed by IBM might be appropriate.[14]The first meeting between IBM (Messrs Thomson, Burgess and Keith) and Southwark (Messrs Katz and Orrom amongst others) took place on 28 April 2006. Mr Thomson delivered a slide presentation, entitled "Integrating Data Throughout the Enterprise". Much of this involved the provision of standard pre-sales information about the Ascential software which indicated that there had been some 5000 global customers for this software including a number of well-known names, such as internationally known banks. It identified duplication as a problem to be addressed and the need for data cleansing and consolidation to be done. It put forward the WIIS and looked at the three tools, Profile, Data and Quality, Stages. It made it clear that QualityStage used "a probabilistic matching engine with customised business rules to deliver the highest quality match rates in the industry”. One of the slides indicated how the equipment could be used to determine that there was a match between two disparate records. The three tools could provide a process comprising “Data Investigation, Standardisation, Matching and Survivorship”. The IBM personnel made it clear that these three software programmes did not on their own or even in tandem comprise a complete MDM system. However what they would do would be the transfer, cleansing and integrating of data as part of an overall MDM system. "Cleansing" of data primarily involves the standardisation of the way in which the information is presented and recorded in the particular record; thus, it might be that standardisation would involve always putting the surname first followed the forename. If all the records were in that form, that would effectively make it much easier for the computer systems to read and then process them. Mr Thomson was impressed with the apparent abilities, enthusiasm and experience of Mr Katz and Mr Orrom who seemed well able to understand what was being presented to them. There was probably some general reference to WCC but no explanation.[15]There was a further meeting on 2 May 2006, one purpose of which was to discuss WCC. Messrs Thomson, Burgess, Keith, Brown and Romano attended for IBM whilst at least Messrs Katz, Orrom and Nuttall attended for Southwark. The IBM team explained in some more detail what WCC could provide and that it was an MDM system in itself. They made it clear that the cost of WCC would be between £1.5 million and £2 million. However the Southwark team indicated that they had a budget only of £500,000 and it would be difficult to get authorisation above that figure. It was then resolved that Southwark would continue to look at and consider the more limited WIIS applications. IBM personnel explained that this suite of tools would first deal with data analysis and then data cleansing. It was at least provisionally considered that Southwark would be responsible for project managing the MDM project and would choose the component parts and build the MDM system, itself choosing the services required as the project progressed. These matters were broadly summarised in an e-mail sent to Southwark by Mr Thomson on 2 May 2006 after the meeting.[16]Mr Thomson had some recollection that in the past Ascential had sold a composite software package system which incorporated the three IBM (formerly Ascential) tools with some Orchard software. He approached Orchard and in particular Mr Leventhal who by e-mail on 15 May put forward Arcindex, and a cost for that software together with the suggestion that 31 consultancy days could be provided including training. He also sent a copy of the "White Paper for Arcindex” which was a standard sales document issued by Orchard which broadly described what it could do. These were passed on to Southwark shortly thereafter[17]On 25 May 2006, Mr Leventhal came to London and initially met Messrs Thomson, Burgess and Keith to show them Arcindex which, as was the case, he believed that none of them had ever seen before. They gave him the SAP Brief, which he considered. He believed that Arcindex would provide what he termed a “very good fit”. All four then went on to a meeting at Southwark’s offices with Messrs Katz and Orrom. Mr Leventhal demonstrated Arcindex to them by reference to a PowerPoint presentation and then by a demonstration of Arcindex on his lap top, which entailed the standard product walk-through showing standard functionality (as described in the Arcindex User Guide which was provided). His evidence, that he described the matching and cleansing concepts and operations within Arcindex as well as the process for identifying and resolving manual review items, was unchallenged and I accept it. Messrs Katz and Orrom gave every impression that they really understood what they were being told and shown. They indicated that as demonstrated Arcindex was the sort of thing they wanted. Their focus at the meeting was on having the ability to find and report where anomalies in the information being processed were rather than having the data changed automatically at the sources, which they accepted would be difficult to achieve. They made it clear that Southwark had "a fairly technical staff". The Southwark team indicated that they wanted Arcindex to match, de-duplicate and store configurable name and address relationships. Mr Leventhal explained that de-duplication was not a core function of Arcindex but could be achieved within the Websphere tools. He told them that Southwark would need to write SQL (Structured Query Language) scripts to report where duplication occurred and that these reports would then go back to the CRM operators who would need then to remove the duplicates and merge the case history of the duplicated files. He noted that Messrs Katz and Orrom raised several matters (the ability to deal with multiple relationships between names and addresses and to search Arcindex from the CRM system). The Southwark representatives appeared well disposed to Arcindex, saying that it was the best proposal which they had received; however it was agreed that before things moved any further, a pilot demonstration of Arcindex using the actual data would be undertaken.[18]On 26 May 2006 Mr Thomson, Mr Avon and Miss Hampton of IBM met Messrs Katz, Orrom and others to address requirements for the IBM as opposed to Orchard products. As is clear from Mr Thomson’s e-mail of that date to them, it remains unclear whether it wanted to employ IBM to provide its software with a separate procurement by Southwark of Orchard’s. On 30 May 2006, in response to an e-mail query from IBM on 26 May 2006, Mr Katz indicated amongst other things that data from three of the source systems would need de-duplicating.[19]Or about 31 May 2006, Mr Thomson prepared a document entitled "Southwark Council Master Data Manager Estimates". The "Prerequisites" were: “This document addresses Phase 1 of the Southwark Council’s Master Data Manager Project, with respect to IBM services only. IBM’s performance of the Services described in this document is dependent upon the information provided in the following documents. A number of risks and assumptions have been generated as a result of this exercise due to incomplete information being available; these are detailed… Since detailed mapping and design specifications do not yet exist, the estimated timeline … was built from information gathered during a meeting between Southwark Council and IBM on 26th May 2006…” 24. A figure was shown in Paragraph 1.2 which showed the Websphere tools being used in parallel with what was described as the "Orchard MDM". Paragraph 1.3 described the "Scope": “The scope of this project engagement is to:• Profile the data of the identifying sources.• Extract data from the sources.• Cleanse and de-duplicate the data…• Profile the data of the identifying sources.• Extract data from the sources.• Cleanse and de-duplicate the data… This project is phase 1 of a larger project… The scope of this project engagement is NOT to:• Carry out any services pertaining to Orchard software.” 25. "All source systems (apart from Arcindex) will be cleansed and de-duplicated using the following steps:• Investigation - discover patterns within the data.• Matching - identify potential duplicate records based on patterns discovered during the investigation.• Standardization - impose a standard data pattern to take forward.• Survivorship - pass forward with the lowest appropriate data as the surviving data record. One of the express assumptions was that the "success and development and testing is dependent on the accuracy of the Southwark Council detailed mapping specification documents."[20]On 2 June 2006, Mr Leventhal, on his own and without any IBM representation, met Messrs Katz, Orrom and Nuttall at Southwark. He gave a further demonstration of Arcindex and a further PowerPoint presentation. He dealt with the two matters raised earlier, but, with respect to the CRM search, it was agreed that this would require further investigation and discussion, albeit, as it turned out, the project ultimately never got to the stage of dealing with this. Mr Leventhal handed over a functional specification which set out a possible solution to the CRM search requirement. There were direct discussions between them about prices and rates. Mr Orrom told Mr Leventhal that Arcindex had already been put forward as the “solution choice” to Mr Currey and that he was very impressed at Orchard’s approach.[21]On 8 June 2006, Mr Leventhal returned alone to Southwark, meeting Messrs Katz, Orrom and Nuttall again and gave the agreed pilot demonstration of Arcindex using Southwark’s own data consisting of 1000 customers and addresses taken from the LLPG, CRM and EReg source systems, which Orchard had been working on over the previous few days; the data had been standardised in fact using QualityStage. Mr Leventhal demonstrated how data was matched into Arcindex and what happened when data failed to match; this involved showing the manual review process. He demonstrated various Arcindex matching reports and on-screen statistics and also how the index was to be searched for names and addresses; he showed how the audit screens could be viewed to show the history of data being loaded into the index. The demonstration was a "hands on" one with the Southwark representatives physically participating. The Southwark representatives indicated that they had a positive reaction. The following day, Mr Leventhal sent a report on the pilot data load demonstrating how the different data sets had been matched together. On 9 June 2006, Mr Leventhal sent them a PowerPoint presentation showing how Arcindex was being used at North Lincolnshire Council, which had been mentioned previously during earlier discussions. Over this period there were a number of direct communications between Orchard and Southwark[22]By mid-June 2006 there were two schools of thought within Southwark in relation to whether the Council should select the SAP system or the Websphere/Arcindex approach. Messrs Katz and Orrom were firmly in the camp which favoured the latter. In conjunction with Mr Thomson of IBM, these two gentlemen put together a document entitled “IBM Ascential Business Case for Enterprise Data Profiling, Cleansing and Transformation" dated 14 June 2006. Although it was on IBM paper it was a jointly drafted effort, called for by Messrs Katz and Orrom. The document primarily described the Websphere tools and contained a "business case template" which compared the functionality of “IBM & Orchard”, “SAP” and “IBM Ascential”. It showed that SAP was the most expensive. However it showed that for “IBM & Orchard” and "IBM Ascential” a detailed clerical review process to review matches was provided.[23]On or about 22 June 2006 Southwark decided internally that it would not go ahead with the SAP alternative but would in principle go ahead with the Websphere and Orchard solution. Mr Currey telephoned Mr Thompson that evening about the decision. There is now no suggestion that this communication gave rise to any contract between IBM and Southwark. It is clear that there was no contract at this stage because the parties were agreeing that they still needed to discuss, agree and sign off on a contract. What was discussed was that a GCAT contract was envisaged, which was a form of contract envisaged for use where, as here, IBM was a government framework contractor. Later, Mr Thomson was asked to discuss contractual matters with Mr Lloyd and on 27 June 2006 sent through a draft contract based on the GCAT forms. During discussions between the two of them about the contract, Mr Thomson made it absolutely clear that the Websphere package was not "a bespoke solution" but was "packaged software". It is clear that the parties proceeded on the basis that there were to be separate contracts for the Websphere software, consultancy services and Orchard Arcindex. The Contract for the Websphere software was signed on 29 June 2006 (“the Websphere Contract”).[24]Over the following few months, little seems to have happened so far as the involvement of IBM was concerned. Internally Pearson produced a "Project Management Proposal” for the "LBS Master Data Management System". This document at Paragraph 2.1 presupposed that Southwark had contracted with IBM and Orchard respectively for Websphere and Arcindex. There was only one contract in place at this stage, namely between Southwark and IBM for Websphere. This Proposal includes detailed recommendations as to how, with the assistance of independent consultants such as Pearson and Serco, the project could be sensibly managed from Southwark’s standpoint. It called for a "Customer Requirements Document" which would "outline in precise and explicit language the functions and capabilities for the MDM system as well as state any required constraints by which the system must abide." This was never produced by Southwark.[25]There was some discussion and communication between Southwark and IBM in late August 2006 about a “Statement of Work for Assistance Services” essentially relating to time and services to be provided by IBM. This was broadly approved by Southwark and resulted in a second contact ("The Consultancy Services Contract”) being entered into in October 2006. Although nothing now turns on this contract, the agreed Order for the work was prefaced as follows:
“IBM will provide the following assistance services… under this Order to help you implement phase 1 of your project known as the "Master Data Management (MDM) project" which aims to prepare an application that is able to provide a master cross reference index in order to give you a joined up view of person and property across your 5 [source] systems…”
The services included the provision of a technical lead architect, ProfileStage and DataStage consultants and an implementation consultant to provide support. A total number of some 688 hours work was provided for at agreed hourly rates.[26]Mr Leventhal also produced a document outlining Orchard’s statement of work in early October 2006 which, amongst other things, informed Southwark that Orchard would oversee the implementation of the data configuration, cleansing and matching through site visits and remote office based support.[27]A “Kick-Off” meeting was held on 2 November 2006 at Southwark. Mr Orrom, described as the Enterprise Data Warehouse (“EDW”) Project Manager, Mr Ojejinmi the Programme Manager, Mr Huerta the MDM project manager, Mr Hatia from SAP a project manager, a Ms Dungate as Programme coordinator, Ms Robertson as Serco Project Manager and Mr Katz the "Data Architect/Manager” attending on the Southwark side and Messrs Lloyd and Campbell for IBM. At around this time IBM invited Southwark representatives to attend for ProfileStage and DataStage training. Another meeting followed on 14 November 2006 with most of the same attendees at which there was a discussion about hardware which would need to be provided.[28]By late November 2006 there was some concern within Southwark that Orchard was not sufficiently involved. Mr Ojejinmi wrote to Messrs Orrom, Katz, Serco, Leventhal and Lloyd on 21 November 2006 as follows:
“We are moving swiftly ahead with the implementation of our MDM project but I am concerned about the non-involvement of Orchard information systems, especially as we are close to procuring hardware for the Accential [sic] part of the implementation. Having had conversations with Serco and Orchard, it has been made obvious that we need to engage with Orchard ASAP. I have tentatively agreed to have Simon [Leventhal] come down for technical discussion on data and sizing end of this week… James [Orrom], please arrange and confirm and make sure all the relevant people at present. Serco will not be going ahead with procurement of any hardware until we have a complete picture of the whole infrastructure. Also, could we confirm the current state of Orchard contract? Where is the hold-up? Southwark”
Also, could we confirm the current state of Orchard contract? Where is the hold-up? Southwark” This tangentially supports other evidence, which I accept, to the effect that serious consideration was being given to there being a separate contract for Arcindex between Southwark and Orchard. Indeed a detailed draft was produced.

Also, could we confirm the current state of Orchard contract?

[29]IBM produced in early December 2006 for Southwark a "Skeleton Scoping Document" which on its face may have been designed to be completed by Southwark. It was meant to deal with what it termed “data migration” to transfer all necessary data from Southwark’s current applications to the proposed new systems. It contained detailed descriptions of the three Websphere tools. In relation to QualityStage, it explained that it had an "easy to use GUI” (Graphical User Interface). It also summarised all the investigation, standardisation and matching capabilities of QualityStage; these included by way of bullet points “Facilitate effective matching", "More matches, less errors, less clerical review” and "No limit to the number of match fields”. It describes the benefits of QualityStage as including "the most powerful, accurate matching available - based on probabilistic matching technology and a full spectrum of fuzzy matching capabilities that are easy to set up and maintain and provide the highest match rates available in the market" and "an easy to use GUI with an intuitive, point-and-click interface for specifying automated data quality processes-data in investigation, standardisation, matching, and survivorship."[30]In or by early December 2006, Alison Leahy was becoming involved in the project on Southwark’s side. She was, initially, the Information and Knowledge Manager. Mr Orrom e-mailed her on 30 November 2006 to explain what was involved in the overall Enterprise Data Warehousing project: “SAP Business Warehouse - Provides data warehouse technology. Info cubes, query tools, analytics SAP Portal - Provides the presentation layer for the web based reports/analytics SAP Xi - Provides the system integration between data sources and SAP BW IBM Profile Stage - Provides the data analysis tool. Assesses data quality and looks for suitable fields to match against IBM Data Stage - Provides data cleansing functionality Orchard arcindex - Provides the Master Data Repository. This is a separate data base that matches the duplicated records together and holds them against a master "golden" copy." SAP Portal - Provides the presentation layer for the web based reports/analytics SAP Xi - Provides the system integration between data sources and SAP BW

IBM Data Stage - Provides data cleansing functionality

[31]Mr Leventhal of Orchard discussed directly with Southwark what was needed to install the Orchard equipment; for instance on 8 December 2006 he e-mailed Serco and the Southwark team with information on this. On 8 December 2006, Mr Campbell of IBM provided some advice to Southwark about how information could be extracted from the CRM source system and also sent standard answers to frequently asked questions about the IBM software.[32]By mid-December 2006, it had been accepted that the contract for the Orchard software and services would be between Southwark and IBM. The main reason for this seems to have been that Southwark would thus avoid having to go through a public procurement exercise. On 21 December 2006 Mr Thomson sent to Mr Orrom a draft of the contract. It is not clear why it took some two months to secure a final agreement thereafter. Some contractual arrangement was made as between IBM and Orchard at some stage but, other than the generality, there is little evidence about precisely what it was; and there is in the papers an Amendment Letter dated 23 February which suggests that the IBM agreement with Orchard was referenced 4906UK1557.[33]Southwark and Pearson produced in early January 2007 a "system integration model" which figuratively on paper represented what the MDM scheme overall would provide. It shows information being fed from the source systems through DataStage and ProfileStage and then to QualityStage and Arcindex. From there, “data exceptions” could be sent back through an "exception reporting process” which would correct errors in the data and the "cleansed" detail would then revert back to the source systems. There was to be a "Portal (View)" of the QualityStage/Arcindex area of the system which could be accessed by the "Source System Owners” who were Southwark and Pearson; this would enable access to QualityStage and Arcindex. IBM was wholly aware of this model.[34]As from 26 January 2007, Mr Katz, who had become the Southwark MDM project manager, internally issued the first of his weekly reports which highlighted progress, risks and issues together with tasks and milestones due for completion. On 13 January 2007, Southwark raised a Purchase Order to IBM for the supply of the Arcindex software and related consultancy services. There was a detailed discussion on both sides about the contract for this software. In Mr Katz’s weekly report for 2 February 2007, he indicated that the "Orchard Contract" in its final draft was prepared for review and sign-off. On 31 January 2007, Mr Katz was writing to Serco (not copied to IBM) asking for there to be a review of the Arcindex technical specifications. Somewhat surprisingly, Serco responded on 2 February 2007 to the effect that it had no information or documentation on Arcindex albeit no problem was anticipated. At about this time, there was an unsuccessful attempt by IBM to install the Websphere software which failed because there were problems with the hardware which Southwark had purchased. There has been no suggestion that this was the fault of IBM[35]On or about 9 February 2007 Southwark and IBM signed the third and final contract ("the Arcindex Contract") relating to the Orchard software. This contract was the subject matter of extensive negotiated changes in the various terms which were put forward. At no time, prior to the Arcindex Contract or prior to the other two Contracts, did any representative of IBM make any recommendation to Southwark about the suitability or desirability of using Orchard or Arcindex; there has been no evidence from Southwark that there was any such recommendation; the documents do not suggest the same and IBM witnesses gave unchallenged evidence, which I accept, that no such recommendations were made. After this time, Mr Katz continued to make direct contact with Mr Leventhal at Orchard, often without copying in IBM. It is also correct to say that there were a number of meetings and contacts over the next three months between IBM and the Southwark team members. By this time Ms Leahy, who did not have a particular background in IT was put in charge of the MDM project as Project Executive.[36]In his report of 2 March 2007, Mr Katz reported that "IBM Resource back on board and delivering benefits" and that "Customer Service Delivery Issues Resolved with IBM Senior Management". These matters, such as they were, had been discussed on or about 28 February 2007 between Southwark and IBM. On 9 March 2007, Mr Katz and a Pearson (now Vangent) representative produced a Business Requirements Specification which at some stage was distributed to IBM. It indicated, amongst other things, that within Southwark "access to the MDM master repository [Arcindex] will be role-based and determined on business need”. This is relevant to one of the complaints relating to access. It repeated, virtually verbatim, some of the bullet points about QualityStage referred to above (such as "an easy-to-use GUI” and "the most powerful, accurate matching tool available”).[37]By 16 March 2007, as Mr Katz reported, IBM had completed the installation of the Websphere software. In his report, he described this as one of the "key successes" that week. There continued over the next few weeks to be close liaison directly between Mr Leventhal and Southwark, in particular Mr Katz about the impending installation of Arcindex. Thus, for instance he sent a document entitled “arcIndex Training Schedule and Prerequisites" on 21 March 2007. On the same date, Mr Kronfeld of IBM reminded Mr Katz of the understanding that a “mapping specification" was required from Southwark and indeed this was previously agreed and recorded on a project plan. He actually provided a document which had to be filled in and completed by Southwark. This mapping specification was never produced by Southwark and was the primary reason that IBM could not complete its Websphere installation. No good or indeed any explanation has been provided as to why Southwark did not produce the mapping specification which it had undertaken to provide.[38]On 17 April 2007 there was what was termed a Project Board Meeting attended by Southwark (Ms Leahy, Mr Katz, Mr Cottrell), IBM (Mr Avon), Orchard (Mr Leventhal) and representatives of Vangent and Serco. Ms Leahy said that there was a lack of a single strategy or vision for the whole project beyond the first phase; this needed to be defined. It was recorded that the "project lacked the necessary teeth to be able to effect the appropriate technological and business changes needed for the success of the project". There should be a requirement that "data cleansing" was adopted by the source system owners. Mr Leventhal again demonstrated the Arcindex software without demur from anyone present.[39]By 27 April 2007, Orchard had completed the installation of the Arcindex and related software which was described by Mr Katz in his report of that date as having been "100% successful". By this stage, Southwark was looking to recruit a "Data Integrity Manager" which by 18 May 2007 was being described by Mr Katz in his weekly report as the "key" risk or issue.[40]Towards the end of May or in early June 2007, Mr Avon, the IBM Services Manager allocated to this job, became increasingly concerned that Southwark had still not produced the mapping specification without which IBM could not complete its part of the project. Little work was being done in June 2007 by IBM or Orchard and the project was in delay. Whilst the Websphere software could find and improve the quality of data, it was a tool set and would not automatically provide data in the format wanted. Therefore Southwark had to define what mapping and formatting rules it wanted. He spoke with Mr Katz about these problems and said that, because "IBM was burning Southwark’s money” by having people allocated to the project when they were of little further use until the mapping specification was provided, he would take people off the job temporarily. He told Mr Katz that there was no point in having Mr Kronfeld of IBM allocated to the job continually because he had got as far with the implementation as he could. The Websphere software had been installed correctly, connectivity had been established with the source systems and indeed some data had been accessed, extracted and profiled to demonstrate this. Mr Katz agreed that IBM should stop work and re-engage when the mapping specification was completed and provided by Southwark. Effectively, thereafter IBM did no further substantive work although contact was made.[41]By 8 June 2007, Mr Katz was reporting that the "Data Integrity Manager Position" was filled and the person, who turned out to be Clare Troy, was due to start on 22 June 2007. He reported that this Manager would have to complete work packages because they impacted on “project deliverables”.[42]On 22 June 2007, Ms Troy started work and it is fair to say that her involvement was the cause or catalyst for the abandonment of the project. She was retained as a consultant data integrity specialist particularly because she had project managed a successful implementation of an MDM solution at the Borough of Lambeth albeit using different software (“MultiVue”). Ms Troy was told by Ms Leahy that she had done a review in April 2007 and had developed significant concerns about the type of IT expertise available within Southwark. As Ms Leahy was not called as a witness, and as there is no hint the contemporaneous documentation about it, I do not accept that there was any such review but she told Ms Troy that she had done one. Ms Leahy and Mr Katz obviously thought that they needed a data integrity specialist, however, and that was why Ms Troy was retained.[43]For the first 3 months of her time at Southwark, Ms Troy concentrated on considering Arcindex. She did not look at the three Contracts with IBM and does not appear to have studied in any detail the copious files generated by this project up to June 2007. Indeed, she did not even know until about 1 August 2007 what had been purchased from IBM (as appeared in an e-mail of that date from her to Ms Leahy). She formed at a very early stage an unfavourable view about Arcindex, writing to Mr Katz by e-mail on 29 June 2007 as follows: “Just writing a few notes down on Orchard, know we have to live with it, but the following are things we need to be aware of…in summary it seems to have good matching capabilities (although will need to throw a few test files at it) but its interface for managing the process of matching is really poor, so are its current integration facilities. Good points 1) Matching rules, seem to be pretty dynamic, supports date range checking, nickname etc 2) Allows a set of matching rules to be set up per data source 3) Supports the concept of having separate indexes for business, residents, properties etc 4) Keeps historical address links, so you can see where a person has resided over time Limitations 1) Interface is pretty poor, does not support the resolution of possible matches and has limited support when manual linking records - need to get this one sorted out as our data matches will not be able to do very much 2) Will have to run matching for a file several times to get all the duplicates for a record, basically matching stops for a record once it's found one exact match, ideally you want to find all matches for a record in one hit 3) Due to the above will miss potentially data quality issues 4) No lower threshold to discard matches, this is a pain when processing matches and ascertain how many records you really have to process. It will be a problem for the next phase when using the index to search for people from CRM as more records [than] necessary will be passed back 5) Has no reporting capabilities 6) Composition of golden records, is good from the point of view it allows definition of the most trusted data source down to attribute level but does not take into account the age of the record e.g. a record which reflects the most recent contact a person has had with the council is usually going to be the most current 7) Doesn’t currently support integration with other systems Do you have the ERD [Entity Relationship Diagram] for Arcindex? Anyway was a very worthwhile session, but I would say other products out there have the edge on them, including dare I say MultiVue. Tony [of Brent LBC] has invited [us] up to Brent for a demo so suggest we take him up on that so you'll get an appreciation as to what we need to head for." 1) Matching rules, seem to be pretty dynamic, supports date range checking, nickname etc 2) Allows a set of matching rules to be set up per data source 3) Supports the concept of having separate indexes for business, residents, properties etc 4) Keeps historical address links, so you can see where a person has resided over time 1) Interface is pretty poor, does not support the resolution of possible matches and has limited support when manual linking records - need to get this one sorted out as our data matches will not be able to do very much 2) Will have to run matching for a file several times to get all the duplicates for a record, basically matching stops for a record once it's found one exact match, ideally you want to find all matches for a record in one hit 3) Due to the above will miss potentially data quality issues 4) No lower threshold to discard matches, this is a pain when processing matches and ascertain how many records you really have to process. It will be a problem for the next phase when using the index to search for people from CRM as more records [than] necessary will be passed back 5) Has no reporting capabilities 6) Composition of golden records, is good from the point of view it allows definition of the most trusted data source down to attribute level but does not take into account the age of the record e.g. a record which reflects the most recent contact a person has had with the council is usually going to be the most current 7) Doesn’t currently support integration with other systems MultiVue had been the software used for MDM at Lambeth.[44]She wrote on 2 July 2007 to Mr Ellis at Brent as follows:
“I’m starting to get to grips with what is going on at Southwark and spent Friday with Orchard, the supplier of the matching product they have here, Arcindex. I have to say I was a little disturbed at what I was shown, there is no real user interface for resolving matches, and a very limited one for searching. The matching strategy they employ means you have to do several match passes to identify all the matches for a record, which we both know requires far more manual effort to produce a clean index. Basically the way it works is once an exact match is found for a record in a match run it starts and doesn't look for any further matches, potential or exact. They also have not got any production Web services available yet to support integration…oh no, not that one again. On the plus side, their matching rules seem to be far more flexible than MultiVue’s, they can do date range matches, nicknames etc, although I've not had chance yet to put it to a test to see how well it does the job. They have used IBM, datastage as the basis for their matching engine, so in theory it should work. I never thought I'd say this, but based on what I've seen so far, MultiVue, at present, is a more advanced product. Anyway, I would like to take your offer up and come for a visit. It would be really useful to show Alison & Mike what a proper matching product looks like -I’m assuming Initiate has a decent UI [User Interface] to enable the resolution of matches? Given you and Mike are both on crutches, when would be a good time to come and visit you?”
As this e-mail makes clear, Ms Troy had had a meeting at the end of June 2007 with Mr Leventhal of Orchard at which he explained what Arcindex could do. In relation to the proposed meeting with Brent, Mr Ojejinmi was clearly annoyed about Mr Katz being asked to attend at Brent and he wrote to Ms Leahy expressing his concern on 5 July 2007. He was concerned saying: "If the MDM project is looking to head in another direction, then you have to let me know".[45]One relatively minor issue is whether Ms Troy came to Southwark in some way pre-disposed against Arcindex. I have formed the view that she was not as such predisposed against Arcindex, which she had not come across before. However, her experience was with MultiVue used at Lambeth and she was on good terms with Mr Ellis of Brent which used another MDM type product, Initiate. Her comparisons were therefore with two other systems which may well have been procured on a different basis and with different expressed requirements. Certain it is, for one reason or another she became disenchanted with the choice which Southwark had made in relation to Arcindex. She certainly had established views as to what an MDM system should provide.[46]Ms Leahy was influenced by Ms Troy’s advice but she asked Mr Katz on 5 July 2007 for information about procurement and what precisely were Southwark’s requirements. He replied that, having checked the documentation there were "no documents that pertain to the procurement process and the functional requirements/scoring of Southwark’s MDM solution”, but only documents in which he was aware related "to commercial contracts and related budgetary information”. Mr Katz reported on 6 July 2007 that Ms Troy was undertaking a "draft due diligence report on MDM Orchard Software". On 9 July 2007, Ms Troy wrote to Ms Leahy and Mr Katz:
“Just to confirm, the current short term plan plan [sic] for the ‘keep Orchard and enhancement it’ option is 1) Alison to supply strategy/Vision statement for MDM going forward beyond phase 1, to enable us to consider how we want to use Orchard in the longer term 2) Clare to produce a list of the evaluation criteria by end of tomorrow to give to Orchard, taking into account know [sic] gaps, Alison vision and Brent tender document. Objective being to use this document as a basis to identify gaps in the product, cost out these gaps in terms of Orchard Cost, our time and how long it will take to deliver and hence gain a better understanding as to how much it will cost to get Arcindex to do what we want to do. 3) Meeting with Orchard on 16th July, to start the above gap assessment process 4) Agree next steps and timeframe We will need to cost the other options as well…”
We will need to cost the other options as well…”[47]Ms Leahy in an e-mail dated 9 July 2007 to Ms Troy said that a "strategic view for MDM has not been defined" and she provided some bullet points as to what might be the initial and ultimate strategies for MDM. Ms Troy sent to Mr Leventhal on 10 July 2007 "a draft document of the functionality we would require of a ‘perfect’ matching product in areas we need to cover to assess where the gaps are". This comprised a 10 page document which listed requirements and questions which called for a response. Two of the requirements were:
“View old matches generated from a particular match run: exact, potential and non-matches with the ability to resolve potential matches on screen by the click of a button. Full details of the two records performing the match must be available to allow the user to make a judgement as to whether to match the records or not without navigating away to another screen. Also require to be able to enter notes against a potential match for cases where it cannot be immediately resolved… Resolving Matches For a match run need to be able to view: -exact matches i.e. matches that score above a threshold -possible match clients/business -unmatched clients/businesses”
Resolving Matches -exact matches i.e. matches that score above a threshold -possible match clients/business -unmatched clients/businesses” 54. Mr Katz received this document responded to Ms Troy on 11 July 2007:
“…One thing to mention is that a lot of the functionality etc is delivered in combination with Orchard and IBM software, so where Orchard may not have the capability, the requirement is delivered by IBM! If you have a minute I can explain more about this…”
If you have a minute I can explain more about this…” It does not appear that Ms Troy took him up on this offer.[48]At Mr Leventhal’s suggestion, Ms Troy contacted a Mr Foston of North Lincolnshire County Council which was using Arcindex. As set out in her e-mail to Ms Leahy on 13 July 2007, she was surprised that he "was fairly positive about Arcindex” but thought that might be "because he hasn't seen anything else" and because the population demographics were different.[49]On 16 July 2007, Mr Leventhal responded to some of the requirements and questions. Alongside the first of the two points set out above (in Paragraph 47), he said that further development could be provided "to store all candidates from a match as possible matches in the list". He also said that one could "add ability to ignore/bypass all candidates matches below a certain match weight for each class/block”. On the same date Ms Leahy wrote to Ms Troy and Mr Katz saying that, based on Ms Troy’s evaluation, Arcindex “does not currently deliver some key functionality” which Ms Troy felt that it should have. Tasks were listed which included Ms Troy familiarising herself with the IBM software and producing and actioning Arcindex evaluation and specification documents with Orchard providing cost and time estimates for any further Arcindex developments.[50]On 19 July 2007, with little consultation within Southwark, Ms Leahy decided to close down the MDM project. She informed the project manager, Mr Katz, and Mr Leventhal on 19 July 2007; they clearly had not been consulted. Mr Katz was particularly upset, responding that she "could have waited a day until we had spoke more about this” and asking how she expected him to "handle queries where I had been left in the dark about a project I am meant to be managing!” Ms Troy in evidence said that she did not consider that the project was being closed down, but merely suspended. There are however numerous references in e-mail documentation in July and August and September 2007 to the project being closed down and I am satisfied that the Project Executive, Ms Leahy for better or for worse did close down the project. Much of what followed must be viewed in this light; it effectively involved the securing of information which would explain or possibly justify what had or had not been done by Southwark, IBM and Orchard. Mr Leventhal was taken aback at the decision because he considered that the project had not got fully under way at that stage. There remained a significant amount of data analysis and cleansing to do and the profiling, standardising and cleansing process had not yet started because Southwark had not applied the necessary resources to that end. He felt that Southwark were completely re-defining their requirements and, as far as he was concerned, the standpoint being adopted by Southwark ignored the discussions, demonstrations and pilot which had been provided before any contract was entered into and also the User Guide and White Paper for Arcindex. Several days later, he provided contact details to Southwark of another Arcindex customer, the Ordnance Survey, which Southwark felt unnecessary to take up. By this stage there still remained a substantial amount of hours’ and days’ worth of time which IBM and Orchard had not yet provided pursuant to the Contracts. Indeed, these were never called for again by Southwark.[51]Ms Leahy also informed IBM, Orchard, Pearson and Serco of her decision on 19 July 2007. She said that the reason was that the Arcindex application did not "deliver required functionality". It is perhaps of interest that she did not blame IBM at this stage.[52]Over the next 3 to 4 weeks, Ms Troy liaised with Mr Leventhal and numerous e-mails were exchanged as she tried to get to grips with Arcindex and the functionality which it did deliver. She was in the process of producing a "functional requirements document". For instance on 26 July 2007, she asked whether Arcindex satisfied eight listed requirements in her e-mail to him including whether it could "accept data in real-time from source systems or file batches if required”. This was to become an important requirement, albeit one about which no complaint is currently maintained. He responded on the same day. By the end of July 2007, at least internally, Ms Troy had produced a functional and technical requirements document which identified what requirements Arcindex did or did not provide. By this time it was being resolved that the installed Arcindex should be tested using raw, un-cleansed data from the CRM source system. Mr Leventhal agreed to this because he wanted to demonstrate the functionality and capability of Arcindex to Southwark.[53]On 2 August 2007, Ms Leahy e-mailed Mr Avon at IBM asking for basic details about the Websphere tools. The basic question was what each of them contributed "to the MDM solution, individually and in concert”. This suggests strongly that she had not spoken either to Mr Katz or Mr Orrom or looked at the various documents prepared in whole or in part by IBM in 2006 which explained broadly what each of them could do. It also suggests that she had no understanding of what they were for.

Resolving Matches

[54]By 3 August 2007, Ms Troy had produced for internal consumption a more final version of the document with listed functional and technical requirements and whether they had been met by Arcindex and also a "draft functional specification" which was to elaborate on what she considered to be "the missing requirements". This latter document stated:
"The Arcindex product in its current form does not adequately support the MDM function that LBS is planning to put in place, both from an operational and technical point of view. At the time of purchase, it was known that Arcindex did not support integration to other applications and some budget was put aside to enable this to occur in the second phase of the project. The product was recently reviewed and was found to be deficient in a number of other areas…namely the user interface, which seriously impacts the implementation of the first phase of the MDM project. The tool is currently focused more towards the technical user, with a strong knowledge of SQL and is not designed to support the manual resolution of a high volume of potential matches which inevitably there will be with LBS records. The approach to identifying matches is also of some concern, the matching cycle for a record stops once an exact match is found rather than continuing to find all the possible matches for a record. This [area] will need further analysis and testing to fully assess the gravity of the issue.” She identified certain mandatory application requirements which included "potential matches" and a mandatory integration requirement for the products to be able "to accept records in real time from source systems to maintain the index concerned"
. There were a number of other requirements, the large bulk of which do not feature as complaints in the current case.[55]Discussions within Southwark and between Ms Troy and Mr Leventhal then turned to the consideration of loading the installed Arcindex with data from the CRM source system. On 6 August 2007, he explained by e-mail to her that the task could be carried out and went on: “However, it will require a bit of effort, as we will need to configure the environment around the data and have the procedures in place from IBM to actually extract the data we need, in the format required. I would also like to hold back on this request until we have all had a chance to sit down and discuss "the big picture" - where we are with the project and the logistics around the general development requirements."[56]It was only on 8 August 2007 that for the first time Ms Troy received some indication from Mr Katz what the IBM Websphere software could do. In an e-mail on 9 August 2007 to Ms Leahy, she referred to this and suggested that such software was in any event not needed anymore for the integration of data because some SAP software was now planned to do that exercise. On 10 August 2007, Mr Avon of IBM sent to Ms Leahy and Ms Troy an overview of what the Websphere software could do. Overall, the document said that the software spanned the “full data integration life cycle”, it addressed “data quality and standardisation to identify, correct and reconcile inaccurate or redundant data” and provided "Data transformation, to enrich data from any source and deliver it to any target." It explained that QualityStage provided amongst other things "standardisation and matching rules for subsequent data quality management initiatives".[57]On 14 August 2007, Ms Troy sent to Mr Leventhal the latest version of her "Functional Specification for Arcindex Enhancements", saying that these enhancements were required to Arcindex to enable it to “support the MDM project going forward". There were some 13 mandatory "application", "integration", "reporting" and other requirements which it was said had not been met by Arcindex. These included requirements to identify "potential matches", "construction of golden record" and the "real time" criterion. The large bulk of the remainder of the requirements are no longer pursued as complaints in this case although several of them overlap with the "potential matches" requirement. Ms Troy on a number of occasions described what she was doing as a "short research project to look at our MDM solution".[58]Mr Katz on 10 August 2007 produced for internal consumption his "End of Project Report". It stated amongst other things: “Within Phase 1 of the MDM project will [sic] provide the Council the capability to realize operational benefits through the tasks of data cleaning, standardisation and matching. It must be clearly stated that MDM in Phase 1 delivers a capability to Council that it does not currently have, and that the enablement of this capability will provide benefits within the organisation. The Project was closed down by the Executive Officer on Thursday 19th July before any of the capabilities listed above could be realised… As the Project was prematurely closed down before go-live the end goal objectives were not realised, subsequently it is not possibly [sic] to measure performance against targets." On 21 August 2007, Mr Katz informed everybody involved at a senior level with the project including IBM and Orchard that the MDM Project had now been shut down "in its entirety as of the 3rd August 2007". He went on to express his gratitude to all involved for "the professional manner and enthusiasm you brought to the project before it was shut down prematurely".[59]On 22 August 2007, there occurred the test of Arcindex which had been organised by Ms Troy. Mr Leventhal attended. A file of some of 400,000 records from the CRM source system was provided; this was raw, that is un-cleansed and non-standardised, data, which had not been put through the IBM Websphere software. No matching rules had been considered or set up. Mr Leventhal, when he arrived, was given 30 to 45 minutes to set up the test, which was, in his correct view, clearly insufficient. One of the important things one had to do to produce a meaningful test for Arcindex was to identify "rules" which set criteria by which matches in the records could be identified; this could not be sensibly done in the time available although he set some rules for a basic but non-typical matching strategy. Although more than one pass can and could have been done, only one was done for this test. It was accepted by Ms Troy in evidence that at the time the test as far as it went was a success. She wrote an e-mail on 23 August saying:
“Had a successful day yesterday, we loaded with the majority of records from the extract file, matched it and found a significant number of exact matches i.e. record duplicates. I’m still analysing the results, but should be able to pass a report to you of records requiring attention by the end of next week. As mentioned before, this will by no means be a list of all the duplicates you have, due to current limitation in the matching application and a lack of time and resources, however it should certainly improve your dataset on what it is at present.”
[60]Ms Troy sent some of the results to Mr Leventhal on 24 August 2007 who responded that they could have "achieved most if not all the expected results given some additional time." He went on to say that the "software is flexible enough to cater for virtually all the scenarios provided” but regretted that "unfortunately, we only had (I think) about 30-45 minutes to set the rules up from scratch." On 28 August 2007, he sent a document which identified what the additional development costs would be for the main requirements which Ms Troy had previously indicated were lacking from Arcindex. Thus, for the provision of a "Potential Matches Screen", costs would be £40,000 whilst the "construction of Golden Record" would cost some £5000. These prices were later to be discounted by 30%. None of these suggestions were taken up by Southwark. Some consideration was given internally within Southwark to the results of the tests. A document entitled "CRM Matching Test Results" was produced, which confirmed that 48,819 matches or duplicates were found which Ms Troy accepted was a "good result". Ms Troy was apparently away for the next two weeks and matters were not advanced significantly further. Ms Leahy emailed IBM on 7 September 2007 to say that Southwark had closed down the project "because of significant concerns about solution we have procured".[61]Ms Leahy began a draft of a “Review Findings Report for Master Data Management” for internal consumption by Southwark. This document went through a number of changes but the first draft was prepared by her on 4 September 2007. This draft did not seek to criticise IBM at all. The “Main Findings” in Paragraph 2.3 were:
“Business is engaged and supportive of the MDM initiative, however the proposed MDM programme does not have a specific business sponsor although some individual projects do; Hardware framework is in place to support and MDM environment and the data integrity team required to manage the operational management of MDM has been designed and approved; There are a number of significant gaps in the functionality provided by the existing MDM solution procured to support MDM project phase 1, mainly due to a lack of definition of MDM’s requirements and defining strategic programme of work; MDM’s scope was incomplete and integration of MDM to other systems was not fully considered; Data integration strategy and infra structure required to support the strategy needs to be defined. In light of these findings, it is recommended that LBS procure a new solution ‘fit for’ purpose." The draft review also identified that QualityStage provided by IBM was “the matching engine used by Arcindex to match the loaded data files.”
It also confirmed that IBM “had not been formally consulted by LBS before on the topic of integration”. It suggested, based on patchy local authority experience, that implementing MDM was a complex proposition and many have “tried and failed”. Arcindex was said not to be "a mature product… due to its limited user interface, integration capabilities and matching output”. “Real time” integration was expressed to be "a key component".[62]It is clear from Ms Troy’s email to Ms Leahy of 25 September 2007 that it was not until the day before that the two of them had seen the SAP Brief and that was only because IBM had sent them it. Ms Troy emailed Mr Burgess of IBM on that day by reference to the SAP Brief suggesting that Southwark was "not asking for anything new from that time” but she added "we do have a clearer understanding as to how we want our integration to work". At about this time a "workshop” was set up and the agenda which emerged in late September included determining "if WebSphere tools sold to LBS are appropriate for supporting Southwark’s integration needs”, proposing "a feasible solution that does not involve any further product investment for Southwark and determining "the relevance of retaining QualityStage (matching engine) if Arcindex cannot be made ‘fit for purpose’ without further Southwark investment".[63]The workshop took place at Southwark on 3 October 2007 and was attended by Messrs Burgess and D’Silva and Ms Gibbs of IBM and Ms Troy and Mr Heap of Southwark. No minutes were prepared for circulation but a note was made on the Southwark side for internal consumption. This note materially stated as follows:
“Workshop Findings • Determine if there is a feasible solution to support Southwarks MDM’s requirements with the current Web Sphere modules purchased purchased The current solution will not support realtime integration nor will it support any matching capabilities as QualityStage (the matching engine) requires a front-end to invoke its matching capabilities. In order to support any matching capabilities a new matching tool would need to be purchased or Arcindex would need to be enhanced. Datastage and ProfileStage could be retained to support the extraction of data from source systems in batch to the database of the matching tool… • Clarification of whether Websphere modules are functionally appropriate to support Southwark’s realtime integration needs The version of Datastage implemented only supports batch transactions, in order to support event triggered transactions it will be necessary to purchase a product called CDC at an additional cost…” • Determine if there is a feasible solution to support Southwarks MDM’s requirements with the current Web Sphere modules purchased purchased The current solution will not support realtime integration nor will it support any matching capabilities as QualityStage (the matching engine) requires a front-end to invoke its matching capabilities. In order to support any matching capabilities a new matching tool would need to be purchased or Arcindex would need to be enhanced. Datastage and ProfileStage could be retained to support the extraction of data from source systems in batch to the database of the matching tool… • Clarification of whether Websphere modules are functionally appropriate to support Southwark’s realtime integration needs The version of Datastage implemented only supports batch transactions, in order to support event triggered transactions it will be necessary to purchase a product called CDC at an additional cost…”
[64]On 18 October 2007 following some other e-mail exchanges, Ms Leahy e-mailed Mr Burgess, this being the first time that a complaint was actually made against IBM: “In our meeting on 24/9/7 [sic] we made a presentation to you and your colleagues that demonstrated that the MDM "solution" procured from IBM is not fit for purpose. On that day it was IBM’s assertion that the WebSphere Tools without Arcindex could deliver an MDM "solution". The aim of the workshop on 3/10/7 was to determine if this was the case. The outcome of the workshop was that the "solution" procured from IBM is not fit for purpose. I assert that IBM, with a global reputation, has been negligent in selling us a system that does not deliver the most basic objective of any local authority MDM requirement - a customer index. Southwark Council, having a low initial capability in this area, has taken some time and invested a significant amount of Council Taxpayer’s money to discover this. IBM, by selling us this "solution, is complicit in this situation. To summarise we have these modules/applications that are not fit for purpose; WebSphere… £219,586 Arc Index… £67,000 Additional: Support costs £39,526 + £20,000 pa Consultancy days £68,290 As I understand it, for the last two weeks you have been in discussion with Clare about what IBM are going to do to address this situation, now that the workshop has demonstrated that the original solution is not fit for purpose. My interpretation of the initial options put forward, but as yet not evolved, are that Southwark should acquire further IBM or other supplier’s applications in addition to what we have already procured. 1. Is this correct? 2. If this is correct, can you let me know what IBM’s concrete proposal is for us to acquire a solution that is fit for purpose that does not entail any additional expenditure of council funds? We require the proposal to be SOA-compliant. As you are well briefed on our position and our situation, we will not be attending any further meetings unless you can demonstrate they are necessary to bring this situation to a conclusion. Please advise me, by 9 am Monday 22nd, within what time frame you will provide your proposal." WebSphere… £219,586 Arc Index… £67,000 Additional: Support costs £39,526 + £20,000 pa Consultancy days £68,290 Interestingly, the complaint is more that IBM was "complicit” in allowing Southwark to procure something that was not fit for purpose. Mr Burgess indicated on 19 October 2007 that he would be unable to reply within that deadline. It was dealt with as a formal complaint by IBM. On 23 October 2007, Mr Avon of IBM wrote internally that his view was that from a services perspective "we had done what was asked of us".

Consultancy days £68,290

[65]IBM responded on 12 November 2007. A Ms Larter-Whitcher from IBM had considered the comments made by Southwark and indicated that IBM had explored the suitability of the software products. She went on to say: "IBM believes that the IBM ProfileStage, DataStage, QualityStage software proposed and sold to LBS is fit for purpose. The ProfileStage, DataStage, QualityStage software is a suite of tools (known as Information Server) for the development of a Data Integration solution, which is what LBS requested and have specified in the ‘ MDM high-level requirements’ document. The document describes LBS’s requirements for Data Integration on the MDM project and more specifically for batch based data cleansing and movement and integration of data from different systems, which our products can achieve. At the time of purchase LBS chose not to take a total solution/system option due to the cost implications and decided to contract the individual software and services items separately. In addition, LBS chose to project manage the MDM implementation with assistance from the IBM software services team, (contracted separately to the sale of the IBM software), and to date the IBM services contract has had only approximately 50% utilisation. Having taken all the above into account, we believe that Information Server (including ProfileStage, DataStage, QualityStage) remains a valid product set for LBS requirements in addition to being compliant for SOA architecture. I would like to stress that IBM continues to be dedicated to the success of this project, and we are committed to working closely with the Council to assist in its delivery. Our aim is to work closely with LBS to identify next steps which can enable the project to move forwards.”[66]Internally, the response within Southwark to this was that IBM was not mentioning Arcindex and its fitness for purpose and on 12 November 2007 Ms Troy suggested to Ms Leahy that IBM had proposed Orchard and that matters should be handed over to the legal department. There does not appear however to have been any exchanged response back to IBM. Ms Troy and Ms Leahy produced another iteration of their Review Findings Report which dropped from the former version references to there being a “number of significant gaps in the functionality provided by the existing MDM solution procured to support MDM project phase 1, mainly due to a lack of definition of MDM’s requirements and defining strategic programme of work”. It is the case that Mr Katz, the project manager, left Southwark at about this time without having been asked to play any part in the investigation into what had gone wrong; he seems barely to have been consulted. There is no evidence that Mr Orrom or Mr Currey were ever consulted.[67]There appear to have been some further communications between IBM and Southwark about further software which included a WCC solution which was termed “WCC Lite” but those discussions had faltered by January 2008. On 28 January 2008, Southwark’s “Legal and Democratic Services” department wrote to Ms Larter-Whitcher complaining that the "IBM solution" was "not fit for purpose". Seven complaints are identified of which only two are currently pursued together with one further complaint which was not mentioned. It is suggested that IBM misrepresented "its solutions capabilities" to Southwark and the letter concludes:
“In light of the above facts or failure of IBM to resolve this matter and fulfil its obligation under Contractual Agreement with the Council, failure of which is a breach of contract, the Councils right to terminate the Contract and commence legal proceedings against IBM for damages and costs will become exercise and will within the next 7 day [sic]”
A summary of the dispute was attached to the letter running to 3 pages.[68]IBM replied to this on 13 February 2008 saying that the Websphere software satisfied the requirements and that "real Time integration" was not an original stated requirement. IBM said that it was made clear that Southwark had to conduct its own evaluation of Arcindex and that it could not warrant its suitability. IBM said that it had no liability or responsibility in relation to Arcindex. Southwark replied on 13 March 2008 asserting that the SAP Brief was the official project brief and that IBM gave "expert advice” and made recommendations to Southwark in relation to the decision to obtain Arcindex. It concluded by saying that Southwark "relied solely on the judgement and professional skills of IBM to recommend a product that is suitable to the specified requirements of" Southwark. Negligence and misrepresentation was relied upon. The letter concluded:
“The Councils [sic] is therefore terminating the Contract and commencing legal proceedings against IBM in order to recover the costs and damages for Contractual breach within the next 7 working days, unless we receive an adequate response.”
IBM responded on 4 April 2008 challenging this letter in some detail and concluding that there was no legal basis for a claim for damages against IBM in respect of the purchase of Arcindex.[69]It was not until August 2009 that Southwark commenced these proceedings claiming damages at that stage limited to £717,061 which related to the sums paid pursuant to the three Contracts as well as other sums of money said to have been wasted as a result of IBM’s alleged defaults. Surprisingly, given that the legal department of Southwark had been involved as from January 2008, if not before, most of the installed software from Orchard and IBM was deleted from Southwark’s servers with no or no accessible backups. It was only as late as early December 2010 that it was disclosed by Southwark that little or nothing had been retained in relation to this software. This has created difficulties for the experts on both sides in this case. These Proceedings[70]The original Particulars of Claim served by Southwark relied upon misrepresentations made before the end of June 2006, the three Contracts and an "Agreement" (not then properly identified). Implied terms of fitness for purpose, satisfactory quality and the exercise of reasonable skill and care were relied upon as was an implied term of supplying the MDM system "within a reasonable time". The primary complaint was that Arcindex was "seriously deficient” in that possible or potential matches were not capable of being identified. The damages sought were effectively the return of the sums paid to IBM and Orchard together with time and other services provided by Pearson/ Vangent, Serco and Southwark (in-house). After a number of differing drafts were produced by Southwark, it secured leave to amend its Particulars of Claim which were served on 10 September 2010. In addition to the three Contracts, it relied upon an overarching "Main Agreement" made on the 22 June 2006, a collateral warranty to the effect that the Websphere and Arcindex software together would provide a system of merchantable quality and fit for intended purpose, misrepresentations and negligence. A number of additional complaints were added, including that there were two requirements which were not met, in relation to the "construction of a golden record" and that access to screens should be “restricted to a user profile”. The money claim went up to about £2.5 million to claim certain consequential-type losses.[71]By way of Re-Amended Particulars served on 14 February 2011, which was the fifth day of the trial (albeit most of the changes had been presaged in an earlier draft produced on the second or third day of the trial), Southwark substantially changed its pleaded position. It abandoned the overarching Main Agreement. It pleaded reliance on Section 14 of the Sale of Goods Act 1979, not previously pleaded, as well as reliance on the express term of the ICT conditions forming part of the Arcindex Contract of the need for "satisfactory quality". It also abandoned about half of the complaints about the Arcindex software. During final oral submissions and on instructions, Southwark’s Counsel expressly abandoned any reliance on misrepresentation, collateral warranty or negligence and indeed any reliance on any breach of contract other than of the Arcindex Contract. By the timeofthose final submissions, only two of the complaints were left, the rest having been abandoned. The complaints related to the alleged inability of Arcindex to address potential or possible matches and the absence of effective restriction of access to screens.[72]IBM’s Re-Amended Defence so far as remains material in the light of the concessions made by Southwark expressly pleaded that there was no recommendation by IBM of the suitability of Arcindex and that it was Southwark which undertook the relevant evaluation of Arcindex. It was pleaded that the Arcindex Contract was one by which IBM simply acted as a "conduit" for the provision of Arcindex; IBM gave no warranties of any kind and took no responsibility for the performance of Arcindex. It was denied that there were any implied terms and pleaded that the "unsatisfactory quality" term was not on a proper contractual construction a contractual term. It was effectively pleaded that QualityStage provided effective matching capabilities.[73]Practically speaking, the issues between the parties are now relatively limited. Detailed lists of factual, legal and technical issues have been produced but generically the issues are: 1. What, if any, express or implied suitability obligations did IBM have with regard to Arcindex? 2. If and to the extent that there were any such terms, were there any breaches of contract in relation to "potential matches" or lack of restriction of access? Sub-issues are: 3. Does the Sale of Goods Act apply at all? 4. To what extent if at all does the Unfair Contract Terms Act (“ UCTA" ) apply? 5. Causation.

The Contracts

[74]Key background facts in relation to the Websphere and Arcindex Contracts include the following:(a) At no time did any representative of IBM actually recommend to Southwark Orchard in general or Arcindex in particular as suitable. In effect Southwark chose Arcindex and evaluated it for itself, albeit that IBM introduced Orchard as an organisation to Southwark.(b) Southwark personnel were, and gave the impression to IBM that they were, technically qualified to set up, procure and manage the MDM project. In addition they had access not only to other technically qualified people within Southwark but also to other independent software and hardware providers, such as Pearson and Serco.(c) Southwark personnel in particular Messrs Katz, Orrom and Nuttall took it upon themselves to investigate with Orchard and understand what Arcindex was and could provide. Given the presentations, demonstrations, the Arcindex User Guide and the Arcindex White Paper it is clear, and I find, that they knew exactly what Arcindex could and would provide and that they believed that Arcindex met Southwark’s requirements as they perceived them.(d) The SAP Brief, save possibly at the highest level, has not been proved on a balance of probabilities to have represented Southwark’s requirements for the project and software as procured from IBM. The burden of proof is on Southwark in this respect. Whilst the SAP Brief may have represented its requirements in relation to a possible project with SAP, it is clear that no requirements were formulated in writing by Southwark in relation to the IBM procurement. There is no reason to believe that Southwark’s personnel in 2006 or early 2007 suggested that the SAP Brief should be treated as the Brief for IBM. The facts that Mr Katz never complained about any gaps in functionality and that Ms Leahy in the first draft of her Review Findings Report for MDM stated that there were significant gaps in the functionality "mainly due to a lack of definition of MDM’s requirements” support this view.(e) There is, properly, now no complaint that either IBM or Orchard misled Southwark about what Arcindex could or could not do.(f) Southwark took it upon itself to manage the whole MDM project.[75]It is unnecessary to examine in detail the Websphere and the Consultancy Service Contracts because belatedly, by way of Re-Amended Particulars of Claim, no allegations of breach of these two contracts are made or pursued. What was sold pursuant to the Websphere Contract was expressed to be "IBM Websphere Software license and Maintenance charges". The “GCAT terms” were incorporated. Like the Consultancy Services Contract, IBM undertook to "provide the following assistance services…to help you implement phase 1 of your project known as the "Master Data Management (MDM) project" which aims to prepare an application that is able to provide a master cross reference index in order to give you a joined up view of person and property across your 5 [source systems]". It was made clear that the scope did "not include design and implementation of the master cross reference index which will be undertaken by a 3rd party under a separate agreement”. The “3rd party” was in context clearly to be Orchard.[76]By reason of their latest amendments, and indeed the final withdrawing of all other claims by Southwark through their Leading Counsel in closing, Southwark only rely on breaches of express or implied terms of the Arcindex Contract. It is therefore necessary to look at this in some detail. The primary contract document is the Order dated 9 February 2007 and signed by the parties; the order is from Southwark to IBM and is prefaced with these words:
“This Order specifies all the variables necessary for the completion of an ICT Goods and Associated Services contract”
This ICT form of contract was that used for much government procurement and is predicated upon the basis that the "Contractor" (IBM in this case) has been retained by HM Treasury under a Framework Agreement. Under the umbrella of this Framework Agreement, it was specifically envisaged that "public sector bodies within the UK" (which would include Southwark) could enter into agreements with the Contractor and thus procure goods and services at the rates and levels allowed for in the Framework Agreement.[77]Paragraph 1 of the order was entitled in capitals "The Goods and Associated Services” and Paragraph 1.1 was entitled "Ordered Goods":
“IBM will supply the following Third Party Software: Orchard’s arcindex - Person & Property Database under Orchard’s Software Licence Term’s defined in Appendix A”
Orchard’s arcindex - Person & Property Database under Orchard’s Software Licence Term’s defined in Appendix A” 88. The “Ordered Services” were:
“These services will be provided by Orchard and will involve the installation and configuration of the software on the licensed environments. Orchard will provide services to enable implementation of the master database project. Implementation will involve training and transfer of arcindex skills (including data standardisation, matching and administration) to key Southwark personnel with the objective of Southwark being in a position to go live with the software and ultimately be self-sufficient once the transfer of skills is complete. Further details are defined in Appendix C. 89. By Paragraph 1.3, Orchard was to "have an ongoing responsibility to provide arcindex software support in accordance with Orchard’s Support and Maintenance agreement as defined in Appendix B. Paragraph 1.4, entitled "Contractor’s Responsibilities" stated: “IBM will assist you by providing the resources as set out in the Charges section 6.”
Paragraph 4, entitled "Acceptance Test Criteria", provided that the "Ordered Goods will be deemed accepted on delivery".[78]The “Charges” were set out in Paragraph 5. Against "Ordered Goods", the following appeared:
“arcindex one-time charge……£57,000.00”
. In Paragraph 5.2, 70 days’ worth of time was to be charged in effect for Orchard’s services, which included 25 days for "CRM Integration". For "IBM Management" the sum of £8790 was specified, although this was not broken down. A total of £68,290 was allowed against such services. An additional £20,000 was allowable under Paragraph 5.3 for support and maintenance by Orchard.[79]Paragraph 7.1 stated: “In respect of the Orchard software specified in Clause 1.1, the following will apply:7.1.1 Your right to use the software shall be determined solely by the License granted by the vendor to you.7.1.2 No title, ownership, copyright or other intellectual property right is transferred by this Transaction Document." 7.1.1 Your right to use the software shall be determined solely by the License granted by the vendor to you. 7.1.2 No title, ownership, copyright or other intellectual property right is transferred by this Transaction Document." It is, rightly, accepted that the reference to "vendor” in Paragraph 7 is and must be to Orchard.[80]Paragraph 7.2 is an important clause in the context of the issues between the parties:
"Any warranties and indemnities relating to the Program specified in Clause 1.1 are the responsibility of the vendor. The vendor will provide the warranty in Programs in related Services in accordance with the License terms in Appendix A of this Order. There are no additional IBM warranties, and any implied condition or warranty of merchantability or fitness for purpose are excluded.” 81. Above the signatures on the order the following appears: “By signing below you agree that the Order specifies all the information necessary for the completion of an ICT Goods and Associated Services contract and we agree that a contract has been formed using the terms and conditions in the Model contract, augmented by the information provided in the Order, such information being replicated or referenced in that statement.” 82. Appendix A related to Orchard’s licence, relevant parts of which are as follows: “1.1 Orchard hereby grants to the Customer a non-exclusive perpetual licence to use the Software on the terms and conditions contained herein… 6.1 Orchard undertakes that, provided it is operated in accordance with Orchard’s instructions, the Software will be virus free, and will perform in accordance with the specification (Appendix D). Orchard does not guarantee that the Software is free of minor errors not materially affecting such performance. 6.2 The Customer acknowledges that the Software has not been prepared to meet its individual requirements. Orchard shall not be liable for any failure of the Software to provide any facility not specified in the specification. 6.3 Except as specifically set out in this Agreement, Orchard has made no commitment to develop, modify or enhance the Software to meet the individual requirements of the Customer. Where Orchard agrees after the date of this Agreement to carry out any additional software development, including but not limited to any such development identified during the course of implementation as a "go-live" requirement, the specification must be agreed between the Customer and Orchard no later than 20 weeks before the Software is scheduled to "go live"… 8.1 Title, copyright and all other proprietary rights in the Software and the Documentation and all parts and copies thereof shall remain invested in Orchard” The "
Customer" is clearly intended in this context to be Southwark. 83. Appendix B related to the support and maintenance services which Orchard was to provide. Appendix C identified various other services such as "training”. However, under the heading "Pre-Implementation Software Development", there was a reference to what had happened prior to February 2007:
“A key requirement for the Customer is to be able to effectively manage multiple relationships between their properties and population. This was highlighted during initial application demonstrations. This functionality is now available in the Software.” 84. Appendix D although entitled "Application features" is referred to in Paragraph 6.1 of Appendix A of the specification. Although no breach of Appendix D is pleaded, it does identify broadly what is to be provided by Arcindex: “Oracle with Web browser front-end and utilising the Java programming technology. A single view of customers/addresses that exists in many different corporate databases. The single view enables the storage of all back-end reference codes to that person/property in one location. . The ability to bulk-match data from many back-end corporate applications into a master database. . Matching of data by using matched passes that allow sub-setting of potential matches. . Facilities to enable successful matching between datasets that have very different ways of storing data. . Keep a history of when individual data sets were matched to each other-whether this is by way of an initial bulk match into the database or by a periodic update. . Allow non-matching incoming data records to be manually reviewed by an administrator and be manually matched into the database. . The customer can build up extensive matching algorithms by using fuzzy matching and uncertainty word comparisons. Weightings are used to allow a threshold to be set, thus enabling cut-off points to be set that will accept or reject potential matches. . Extensive data searching facilities with access to comprehensive SQL based searching.”
. Extensive data searching facilities with access to comprehensive SQL based searching.”[85]The ICT terms materially were as follows:
“2.1 This Contract governs the overall relationship of the Contractor and a Customer with respect to the supplier of the Ordered Goods and the provision of the Ordered Services. The Customer has ordered the Ordered Goods specified in Schedule 2-2 and the Contractor shall supply those Ordered Goods: 2.1.1 in accordance with the provisions of this Contract if those Ordered Goods are to be sold by the Contractor and purchased by the Customer; and 2.1.2 in accordance with the provisions of this Contract and any other terms and conditions as are agreed between the Customer and the Contract and any third party (if any) if those Ordered Goods are to be leased to the Customer. 2.2 The Contractor shall supply the Ordered Goods and provide Ordered Services in accordance with the provisions of this Contract… 12.1 The Contractor warrants and represents that: 12.1.5 the Ordered Goods (and all components thereof) and the Ordered Services are: of satisfactory quality; in conformance with the relevant specifications set out in this Contract, relevant Order and the manufacturer’s specifications and documentation; and free from material programming errors and material defects in design, manufacture or materials throughout the applicable warranty period, as specified in the catalogue…” 12.3 Except as expressly stated in this Contract, all warranties and conditions, whether express or implied by statute, common law or otherwise (including fitness for purpose) are hereby excluded to the extent permitted by law. 33.2 Neither the Customer nor the Contractor has relied upon any representation or promise except as expressly set out in this Contract. of satisfactory quality; in conformance with the relevant specifications set out in this Contract, relevant Order and the manufacturer’s specifications and documentation; and free from material programming errors and material defects in design, manufacture or materials throughout the applicable warranty period, as specified in the catalogue…”
There were in Clause 13 limitations on what damages or losses could be claimed. Thus for instance, the "aggregate liability of either party for all defaults was not in any event to exceed "the greater of …£100,000 or…120% of the aggregate total of the Charges”. As I have not heard arguments on these clauses and these would go to damages, it is unnecessary to review these. 86. It is first necessary to consider what the material terms of the Arcindex contract were on their proper construction. Given the actual context, it is unsurprising that what IBM undertook to supply was described as "Third Party Software." This was not only because it was in fact Orchard software but it also reflected the fact that it was Southwark which had selected Arcindex as reflecting its project team’s requirements, without any technical or indeed other input from IBM. That is why Clause 7.2 states that "any warranties and indemnities relating to [Arcindex]…are the responsibility of [Orchard]”. IBM did undertake in Clause 7.2 that Orchard would provide the warranty, software and other services in accordance with Appendix A, which is formulated in terms ofOrchard (as opposed to IBM) granting a licence and undertaking compliance with the Appendix D specification, amongst other things. 87. One needs to read Clause 7.2 as a whole and in context. IBM offers no "additional IBM warranties" because it was Orchard software selected by Southwark without input from IBM and because Southwark was contracting on the basis that warranties and indemnities relating to Arcindex are Orchard’s responsibility. There was nothing to stop Southwark from getting direct warranties from Orchard. What IBM was warranting was that Orchard would provide such a warranty, the software and related services in accordance with Appendix A. Other than that, IBM gave no warranties and any implied conditions or warranties as to merchantability or fitness for purpose were agreed to be excluded. 88. However, the effect of the words above the signatures on the last page of the Order makes it clear that the ICT conditions are incorporated, albeit "augmented by the information provided in the order". The use of the word "augmented" suggests that the Order terms increased or supplemented, rather than altered, the standard ICT conditions. If the word "modified" had been used, it might have been easier to suggest that parts of the ICT conditions were to be construed as having been deleted. It follows from this that in my judgement Clause 12.1.5 was and remains part of the Arcindex Contract. However, the three sub-warranties in Clause 12.1.5 (satisfactory quality, conformity with specifications and Order and freedom from error and material defects) have to be read in conjunction with the Order. Thus, satisfactory quality is to be determined by reference to the quality which was called for in the Order, Appendix A and Appendix D. There is no plea that there was non-conformity with specification or that there were errors or material defects. 89. It is argued by Southwark that "satisfactory quality" itself infers at least general suitability for purpose. Analogy is made to Section 14(2) of the Sale of Goods Act 1979, which implies a term where a seller sells goods in the course of a business that the goods supplied are "of satisfactory quality" and to Section 14 (2B):
“For the purposes of this Act, the quality of goods includes their state and condition and the following (among others) are in appropriate cases aspects of the quality of goods— (a) fitness for all the purposes for which goods of the kind in question are commonly supplied, (b) appearance and finish, (c) freedom from minor defects, (d) safety, and (e) durability.” (a) fitness for all the purposes for which goods of the kind in question are commonly supplied, (b) appearance and finish, (c) freedom from minor defects, (d) safety, and (e) durability.”
I do not consider that this definition particularly takes the matter further, particularly in the light of the findings of fact set out below in relation to the alleged breaches. However, even if one was to apply by analogy the substance of this provision, it would still be subject to the overall agreement reached between the parties. Even in Section 14 (2B), the list of matters going to whether something is of satisfactory quality is only to be applied "in appropriate cases"; thus, one needs to look at the contractual context of the satisfactory quality term in question to determine whether and if so to what extent "fitness for purpose" is allowed for. In any event, the "fitness for purpose" element of "satisfactory quality" only relates to the general or common purposes for which goods of the particular kind are supplied. It does not go to specific purposes. 90. Given Clause 12.3 which purports to exclude all other warranties (including fitness for purpose), and subject to UCTA, one then needs to determine from the contract whether there is anything in the expressed requirements in the Arcindex Contract that Arcindex would be suitable for Southwark’s specific or notified purposes. In my judgement, there is no such requirement: (a) Clause 7.2 of the Order makes it clear in effect that the only warranty is in effect that Orchard will warrant as set out in Appendix A. (b) Clause 6.2 of Appendix A contains the clearest expression that Southwark accepts that Arcindex has "not been prepared to meet its individual requirements”. That is comprehensible in the sense that Arcindex was in one sense software bought "off-the-shelf" and in the context that Southwark had wholly satisfied itself that what Arcindex had been described and demonstrated pre-contract to the various Southwark personnel as providing achieved whatever its requirements were. 91. In my judgement, whilst there was a contractual provision within the Arcindex Contract that the Arcindex software would be satisfactory, the satisfactoriness would be determined by compliance, in the context of this case, with Clause 6 of Appendix A to the Order and performance in accordance with Appendix D. If parties to a contract such as this have spelt out what the software must do or be, it will be satisfactory if it achieves what the contract dictates it should. Thus, Clause 6.1 states that the software will be "virus free”; if when operated in accordance with Orchard’s instructions, the software is not virus free, it will not be of "satisfactory quality". If the software does not provide "extensive data searching facilities with access to comprehensive SQL-based searching" as called for in Appendix D, similarly it will not be of satisfactory quality. 92. There is no room in this contract, disregarding any statutorily implied terms, therefore for some interpretation which suggests that IBM was warranting that Arcindex was in some abstract way to be suitable for purpose. If it did not comply with Appendix A or Appendix D, there would be a breach of the "satisfactory quality" warranty incorporated in the Arcindex Contract. The software was to do what these two Appendices said it would. 93. It follows from this that, other than in respect of statutorily implied terms, there is no room, either, for implying terms of reasonable suitability at common law because they would essentially offend against the express terms of the contract, as properly construed. I am also simply not satisfied on the balance of probabilities that any "purposes" were in fact communicated to IBM other than those which were incorporated into the Arcindex Contract. Messrs Katz, Orrom and the other Southwark representatives, who were involved on Southwark’s behalf prior to signing of the Arcindex Contract, ascertained what their requirements were from the Arcindex User Guide and White Paper, together with the demonstrations of Arcindex in mid-2006; by all accounts, they believed and accepted that Arcindex represented exactly what Southwark wanted within the budget which it had. Of course, all implied warranties of fitness for purpose were excluded by the express words of Clause 7.2 of the Order. Statutorily Implied Terms and UCTA 94. Reliance is, ultimately, only based by Southwark on the Sale of Goods Act 1979. It is first necessary to consider whether this Act applies at all in the context of this case. Relevant parts of the Act are as follows:
“1 (1) This Act applies to contracts of sale of goods made on or after (but not to those made before) 1 January 1894. 2 (1) A contract of sale of goods is a contract by which the seller transfers or agrees to transfer the property in goods to the buyer for a money consideration, called the price. 14 (1) Except as provided by this section and section 15 below and subject to any other enactment, there is no implied [term] about the quality or fitness for any particular purpose of goods supplied under a contract of sale. (2) Where the seller sells goods in the course of a business, there is an implied term that the goods supplied under the contract are of satisfactory quality… (2B) [set out earlier in this judgment]. (2C) The term implied by subsection (2) above does not extend to any matter making the quality of goods unsatisfactory— (a) which is specifically drawn to the buyer’s attention before the contract is made, (b) where the buyer examines the goods before the contract is made, which that examination ought to reveal… (3) Where the seller sells goods in the course of a business and the buyer, expressly or by implication, makes known— (a) to the seller, or (b) where the purchase price or part of it is payable by instalments and the goods were previously sold by a credit-broker to the seller, to that credit-broker, any particular purpose for which the goods are being bought, there is an implied [term] that the goods supplied under the contract are reasonably fit for that purpose, whether or not that is a purpose for which such goods are commonly supplied, except where the circumstances show that the buyer does not rely, or that it is unreasonable for him to rely, on the skill or judgment of the seller or credit-broker. 61 (1) In this Act, unless the context or subject matter otherwise requires,— “goods” includes all personal chattels other than things in action and money…” (a) which is specifically drawn to the buyer’s attention before the contract is made, (b) where the buyer examines the goods before the contract is made, which that examination ought to reveal… (a) to the seller, or (b) where the purchase price or part of it is payable by instalments and the goods were previously sold by a credit-broker to the seller, to that credit-broker, “goods” includes all personal chattels other than things in action and money…” 95. A preliminary question to consider is whether the Arcindex Contract was a contract for the sale of goods at all. That involves a consideration first as to whether under Section 2(1) there was to be a "transfer" of "property in goods" and secondly whether "goods" were being sold. I have formed the view that there was here no “transfer” of property in goods for the purposes of the 1979 Act. The Order talks about IBM "supplying" the Arcindex software "under Orchard’s Software Licence Terms defined in Appendix A" and there is a money consideration (£67,000). What was provided by IBM was in effect a licence from Orchard to Southwark to use the software and, therefore, there is no transfer of property. Although the licence is expressed by Appendix A to be a "perpetual licence", this is said to be a licence "on the terms and conditions contained" in Appendix A. Some of these other terms in Appendix A point strongly to there being no transfer of property in the software. Clause 8.1 specifically talks about "title, copyright and all other proprietary rights in the Software” remaining vested in Orchard. Because copyright is identified as a specific right being retained, the use of the words "title" and "other proprietary rights" suggests strongly that ownership rights are retained. The termination provision in Clause 10 requires Southwark to return at Orchard’s request or destroy all copies, forms and parts of the software covered by the licence. Again that points to property in the software remaining all the time with Orchard. The licence was subject to other restrictions; for instance by Clause 3 of Appendix A it could only be used by Southwark and any "commercial partners". The software by Clause 7 was not to be modified without Orchard’s consent. 96. It is therefore unnecessary to decide whether software can be “goods” for the purposes of the 1979 Sale of Goods Act. There is no binding authority which Counsel or I have been able to find on this topic. However, my view, which is necessarily obiter, is that in principle software could be "goods" within the meaning of that Act. I base this view on the following: (a) Although a CD (compact disc) with nothing on it is worth very little, there is no restriction in the Sale of Goods Act on any goods being excluded from the Act by reason of their low value. CDs are physical objects and there is no reason why they should not be considered as goods. (b) The fact that a CD is impressed with electrons to add functions and values to it simply gives a CD a particular attribute. Thus, if a customer buys a music CD with a Beethoven symphony or a Mumford & Sons album on it, it must be "goods” and it should, if new, be of satisfactory quality. There can be no difference if the CD contains software. (c) The definition of "goods" is expressed to be an inclusive rather than an exclusive one. Put another way, the Act is not excluding anything which might properly be considered as goods. It follows that "goods" are not simply "personal chattels", although a music or software CD may also fall into the category of being a personal chattel. 97. So far as this issue is concerned, one will need always to examine the contractual terms pursuant to which the customer acquires the software. If, as here, it is simply a licence to use that is being granted albeit for a money consideration, there may be no transfer of property. In principle, a licence to use, if that is all it is, may well not transfer any property or title in the goods in question. However, if the arrangement between the parties can be said to involve the transfer of property to the buyer, I see no reason why in principle software that is so transferred can not be "goods" for the purposes of the Act. There is in reality only one authority by way of an obiter remark against this view. In St Albans City and District Council v International Computers[1996] 4 All ER 481, Sir Iain Glidewell said: “…in order to answer the question, however, it is necessary to distinguish between the programme and the disc carrying the programme. In both the Sale of Goods Act 1979 section 61 and the Supply of Goods and Services Act personal chattels other than things in action and money…”
Clearly a disc is within this definition. Equally clearly, a programme, of itself, is not. …in the present case, it is clear that the defective programme 2020 was not sold, and it seems probable that it was not hired. The evidence is that in relation to many of the programme releases, an employee of ICL went to St Albans’ premises where the computer was installed taking with him a disc on which the new programme was included, and himself performed the exercise of transferring the programme into the computer. As I have already said, the programme itself is not ‘goods’ within the statutory definition. Thus a transfer of the programme in the way I have described does not, in my view, constitute a transfer of goods." It is unclear in the current case whether the CDs containing the Arcindex programme were to be given to Southwark. Appendix B expressly talks about "new releases" to be "delivered to [Southwark] on CDs”. An inference is that the original software was also provided on CDs. On that basis, the CDs impressed with the software must be capable of being goods, in my view. 98. It follows from the above that the Sale of Goods Act does not apply and therefore the terms statutorily implied under that Act can not be applied to the Arcindex Contract. Even if they did, I am satisfied that by operation of UCTA the exclusions of the statutorily implied terms in the Arcindex Contract would be wholly reasonable and would stand. My reasoning is as follows: (a) I accept that the effect of Section 6 of UCTA is, as between parties such as IBM and Southwark, that the statutorily implied term of fitness for purpose arising under Section 14 of the Sale of Goods Act 1979 will be excluded or restricted pursuant to a contractual term "only in so far as the term satisfies the requirement of reasonableness". (b) Having regard to the matters specified in Schedule 2 to UCTA as Section 11(2) requires, this is clearly a case in which it was wholly reasonable for the parties to legislate as they did. The parties were of broadly equal bargaining strength. Southwark had the opportunity to contract with Orchard and also chose the benefit of standard framework agreement terms. There can be no suggestion that Southwark did not or could reasonably not have known of the terms which excluded the implied terms; Southwark went through a protracted process negotiating the terms and clearly had lawyers involved. Finally, it does appear (as accepted by both experts) that there was some enhancement of Arcindex by Orchard to reflect what Southwark said it wanted. The reality is that the Arcindex contract was not, objectively looked at, an unfair or unreasonable one. By all accounts, Southwark was to get exactly what its staff wanted and was or would have been in the position by Clause 12 of the ICT Conditions to have all breaches remedied, to recover damages, albeit somewhat but not spectacularly limited, and indeed to terminate if there were material defaults which were not remedied within specified times. Was There any Material Breach of the Arcindex Contract? 99. Given the fact that they could not materially access the remnants of the IBM and Orchard Systems, the experts did well to reach as much agreement as they did in their Joint Report No 2 dated 27 January 2011. Material parts of what they agreed in that Report are as follows:
“7. We agree that a major purpose of the project was to improve data quality and remove duplicates. The de-duplication within a single database was a reasonable objective. 8. From the results of the tests that were run at Southwark, we agree that a better mechanism of resolving duplicates would have helped… 10. ProfileStage and DataStage were not used in the test runs done by Southwark in August 2007. 11. ProfileStage would have given information about Southwark’s data and allowed improvement of the matching. We agreed that the matching rules used for the test could have been improved… 12. Some data cleansing could have been conducted in DataStage but not de-duplication, which requires matching of records which is a function of QualityStage… 14. Southwark did not use QualityStage directly for the tests - only through ArcIndex. In other words the user interface to QualityStage was that provided by ArcIndex. Using it this way, Southwark was limited to the functions supported by ArcIndex. 15. As far as we can tell, ArcIndex functions as described in the ArcIndex User Manual and we agree as to how it works and which of the functions relevant to the case it does or does not do. 16. A tool can assist in de-duplication by finding identical or similar records in a database. Whereas identical or near identical can be automatically regarded as duplicates, records that are just similar need to be reviewed before they can be considered as duplicates. It is the latter similar records or potential matches that lead to clerical review. 17. ArcIndex supported a clerical review process to allow the review/resolution of unresolved matches (and unhandled data) but the only way to review potential matches was to read a textual report which could be viewed on screen or printed out. 18. The textual report produced during the test runs in the whole CRM was large but we agree that the amount that needed to be reviewed could have been reduced by manipulation of the rules, for example by setting pass 1 tight (corresponding to definite duplicates/exact matches) and pass 2 loose (corresponding to potential matches). [Note 2 stated that this was "the situation using ArcIndex for de-duplication".] 19. We also agreed that, if the matching rules are set loose to catch potential matches in ArcIndex, it would be necessary to sort out false duplicates before going any further. 20. We also noted that, if the data is of very poor quality, it may not be possible to resolve potential duplicates and this would not be a criticism of the systems.” 100. They also agreed that the SAP Brief is not a detailed requirements specification and that Southwark had not documented detailed requirements prior to February 2007 (Paragraphs 22 and 23). At Paragraphs 25 to 31, the experts accepted that QualityStage had a user interface and that QualityStage could have been used to produce lists of matches, potential matches and non-matches, in report form, an Excel spreadsheet or in a database format. 101. There is no evidence and indeed no suggestion that any of the software, either Websphere or Arcindex, was badly manufactured or installed. The real complaint is that in two respects Arcindex was not fit for purpose. The only two respects ultimately relied upon by Southwark are that there was no effective or proper facility for identifying and dealing with potential matches in the data which Arcindex would address or process and that there was no effective security provision. 102. The answer to all this primarily lies in the determination of what the relevant Southwark personnel identified and communicated to IBM and Orchard as Southwark's requirements for the software (and other services) which they wished to procure from IBM. 103. I have already made it clear that Southwark simply has not proved on a balance of probabilities that the SAP Brief was provided to IBM on the basis that the rather generalised requirements which it contains were and remained Southwark’s requirements for the software and related services project which Southwark embarked on with IBM. This is of some importance because clearly initially Southwark was looking for a complete “MDM System” from SAP prior to IBM’s involvement and even initially when Southwark approached IBM. It was only when Southwark made it clear that they could not afford the £1.5m to £2m price tag which IBM indicated for a complete system that Southwark lowered its sights at least so far as IBM and Orchard were concerned. Whereas the goal remained one of ultimately achieving an MDM system, it seems likely that the requirements for the IBM procurement were significantly less than those for a full system procurement. 104. There is the clearest evidence which I accept that the Southwark team in 2006 carried out a detailed investigation mostly into Arcindex. They received the Arcindex User Guide and White Paper and had at least two if not more detailed demonstrations of what Arcindex could provide. The team was an intelligent one well versed in IT matters and they gave every impression that they fully understood what it was that Arcindex could and would provide. I am led inexorably to the factual conclusion that the Southwark team in 2006 and at all times up to the arrival of Ms Troy was wholly satisfied that Arcindex met its requirements and reflected exactly what it wanted. Southwark has chosen to call as witnesses not one of the five or six people who were intimately involved on behalf at Southwark prior to the advent of Ms Troy. Any one of Messrs Katz, Orrom, Nuttall, Currey and Ojejinmi or Ms Leahy could have provided useful evidence one way or the other about what the communicated pre-contract requirements were but the Court has not been provided with the benefit of their evidence. As I have already said, I do not draw a negative inference as such that they would have given evidence unhelpful to Southwark. However, the evidence from the IBM witnesses and Mr Leventhal points very clearly towards the conclusion referred to above. The very facts that Mr Katz did not record in any contemporaneous documents in 2006 or 2007 that he was in any way dissatisfied with what IBM and Orchard had provided, and that in August 2007 he expressed his gratitude amongst others to IBM for what they had done up to the time that he signed off on the project, particularly in the context of the contents of his End of Project Report, corroborates the conclusion. 105. The experts accept at Paragraph 15 of their joint report that "ArcIndex functions as described in the ArcIndex User Manual”
. As was said colloquially at one stage during the trial, Arcindex does "what it says on the box". An analogy is the potential car purchaser who might want an off-road vehicle but, having looked at the brochure for an on-road vehicle, says to the salesman "that's what I want” and buys that vehicle. There will be no cause of action against the garage that the car is no good off the road. The salesman will reply, with justification:
"you got exactly what you asked for”. That is essentially what has happened in Southwark’s case. 106. The main complaint relating to potential matches stems from Ms Troy’s belief and experience that an MDM system should have a relatively easy to operate method of identifying potential matches. I accept that a primary purpose was de-duplication of the information. That undoubtedly involves finding a way of identifying matches or duplicates so that, for instance, a Mr A Smith who appears on five different records will show up as giving rise to 5 matches, four of which can be identified as duplicates and in effect removed so that there is left one master or golden record for Mr A Smith. It is common ground between the experts, and I accept, that Arcindex should usually be used with data which has already been standardised and cleansed and that then, sometimes by trial and error initially, rules or criteria are fixed in Arcindex for the reception of the cleaned up data. That will produce only two categories: matches and non-matches. There is no separate category as such for potential matches, that is a category of records which do not satisfy the rules or criteria for matches, but which are sufficiently close to merit a clerical review by a human being. However, there was, as the experts said in their Joint Report and indeed in evidence, a number of ways of identifying by way of Arcindex what might fall into the category of "potential matches”. The first is that set out in Paragraph 18 of the Joint Report which involves in effect doing two runs or passes. The first pass would involve the setting of sufficiently tight rules so that only matches or duplicates were found; as Paragraph 19 of the Joint Reports indicates, those matches would then be dealt with so that on the next run or pass they would not show up as matches. The second run would then involve setting looser rules so that a second and different batch of (albeit less certain) matches were identified. These would be potential matches which would need to be the subject matter of a clerical review. Another way of identifying potential matches would be to carry out a review of the textual report either on-screen or as printed out. That review could be tailored so that the reviewers were on the lookout for certain types of closely related information. The names for instance would or could be shown in alphabetical order and that would make a review easier. I accept that this latter approach would have taken substantial human resources at least initially; once the information had removed the duplication from it (the de-duplication initially involving substantial resources), the systems would be rational and only contain single entries for people and places so that later work of de-duplication would involve much less human resource. Of course, the Southwark team recognised that Southwark would have to “cut its cloth” to fit its budgetary constraints. 107. Mr Katz effectively said in his End of Project report that the Websphere and Arcindex software would have provided Southwark with operational benefits from data cleaning, standardisation and matching which Southwark did not currently have and that the project was closed down before any of these benefits could be realised. On any account, Arcindex, if operated sensibly and properly with cleansed and standardised information, would have been of substantial use to Southwark because, at the very least, it could be used to identify and take out a vast amount of duplicated or matching information which would effectively de-clog the source systems and avoid or reduce in a substantial measure the risk of errors being made in the treatment of people and premises. 108. In my judgement, Southwark got by way of Arcindex exactly what its then team knew that they were getting and what it decided that it wanted and needed within its budgetary constraints. It follows from my findings that I am satisfied on a balance of probabilities that Arcindex as supplied by Orchard to Southwark was of satisfactory quality and suitable not only generally but also for the specific purposes which Southwark’s personnel had identified to IBM and Orchard as their requirements. 109. The reality is, now recognised by both experts, that QualityStage was able to provide even what Ms Troy latterly from Southwark’s standpoint believed Arcindex did not provide but should have provided, namely an ability to identify and produce three lists of records, matches, potential matches and non-matches. Subject to one factual conundrum, I would have no doubt that, even if a better provision for potential matches could have been required within Arcindex, it could not be considered to be unsuitable in the context that related software provided by IBM, namely QualityStage, could and would provide that better facility. The factual conundrum relates to the workshop of 3 October 2007 at which there was effectively unchallenged evidence that "there was no feasible solution to support the Claimants’ MDM requirements with the current WebSphere modules purchased” (as set out in Ms Troy's witness statement at Paragraph 86). It remains unclear, even if one accepts this evidence at face value, what was being referred to as the Claimants’ "
MDM requirements". Ms Troy had produced documents listing a large number of requirements which she believed had not been met by Arcindex but should have been so met. Many of those "requirements” have not been pursued in these proceedings, ultimately or at all, as giving rise to any default on the part of IBM. It may well be the case that Websphere could not support all these requirements; for instance, the ability to support "real time integration" is not raised as a complaint against IBM albeit that Websphere may not have been able to provide it; however this was not something which was explored in evidence at the trial. In any event, it was not suggested seriously or at all that in some way IBM was estopped from asserting that QualityStage could and did provide no potential matches function. In reality, therefore in spite of this apparent factual discrepancy, it would not have prevented me from finding that overall Arcindex was suitable in any event and in the light of the availability within QualityStage of the facility to deal with potential matches. 110. Finally, with regard to alleged breaches, I am satisfied that Southwark has failed to prove its case on a balance of probabilities in relation to access to screens being restricted to a user profile. Although it is the case that Arcindex did not support such a restriction, as the experts agree, I am not satisfied that it was a requirement of Southwark. The argument that it was a requirement stems from some general words used in the SAP Brief:
“Definition of user roles and authorizations”
. Quite what was meant by that expression objectively let alone subjectively by the Southwark team involved in 2006 and early 2007 is unclear. It is certainly not a term of art which would have a specific meaning to anyone involved in a MDM system procurement. In my view, all it means is that someone, and it could be Southwark, the supplier of any software or the two of them in tandem, would have to define who would be permitted to use the software and what any given type of or specific user would be required or allowed to do with the software. I do not see how it must be defined as meaning that the software provided by IBM or Orchard has to have built into it an ability to control and restrict access to any given software. 111. In any event, I am not satisfied on a balance of probabilities that, whatever this expression meant or was taken to mean, it was to be a requirement which either Southwark or IBM expected, anticipated or required IBM to satisfy. Indeed I am satisfied on a balance of probabilities that it was not a requirement which IBM was expected to satisfy. My primary reason for making this finding is that the Southwark team knew exactly what they required and what they were getting and therefore must have known prior to the Arcindex Contract that Arcindex would not itself provide restrictions on unauthorised users. The second reason is that there is clear evidence, which I accept, that Southwark was planning on itself providing an access platform to Arcindex through the Portal which Southwark was setting up to enable users of one sort or another to view Arcindex. It should have been possible therefore to secure restrictions on access to Arcindex through the Portal. Overall corroboration for this lies in the (albeit negative) fact that there is no hint in the documentation or evidence that anyone thought that Arcindex should provide this facility until, latterly, Ms Troy came on the scene. 112. It follows from the above that any claim for breach of contract in relation to unsatisfactory quality or lack of fitness for purpose fails and what was left of Southwark’s case as a whole fails on the issue of liability. Causation 113. In the light of my findings on liability, it is unnecessary to consider causation. Out of deference to the arguments put forward however, I will summarise what my views are. Of course, these proceedings are not a public enquiry into what went wrong; these proceedings are concerned with whether Southwark can prove its case on causation as to whether the decision to abandon the software installed by IBM and Orchard was caused by any breaches of contract on the part of IBM. As no breaches have been established, causation in that sense has not been proved. However, there was no evidence from anyone within Southwark involved in the decision-making as to why the installation was abandoned. Ms Troy was a consultant and not a direct employee of Southwark and there is no suggestion that she was party to the decision. In that sense as well, Southwark has not proved its case linking the decision to abandon with any breaches. There appears to have been a growing perception between Ms Troy and Ms Leahy that the MDM project as it had developed by mid June 2007 had been misconceived and inadequately defined. This explains why, extraordinarily, the decision to close down the project was done without reference to the project manager, Mr Katz and why at least initially Ms Leahy and Ms Troy in their Review Findings Report of early September 2007 stated that the proposed MDM programme did not have a specific business sponsor and that there were a number of significant gaps in the functionality “mainly due to a lack of definition of MDM’s requirements”; they went on to say that the scope was incomplete and integration of MDM to other systems was not fully considered. They suggested that in these circumstances what had been provided was not fit for purpose. Southwark later went on to complain that IBM had been "complicit" in effect in letting things go wrong; the use of the word "complicit” is tantamount to saying that IBM should have advised Southwark that Southwark was going wrong. That of course is no part of its pleaded case, as finally maintained. 114. If I had found that Southwark had established the breach in relation to only the restrictions on access, I would positively have found that causation had not been proved. The evidence from the experts was that for a relatively few thousand pounds this problem could have been overcome and the evidence does not suggest that this breach featured significantly if at all in the decision to terminate the relationship between the parties. If I had found that Southwark had established the breach or breaches in relation to the absence of an effective facility within Arcindex to address potential matches, I would have found that Southwark had not proved its case on a balance of probabilities. My concern in those circumstances would have been the absence of any evidence from any decision-maker within Southwark and there would have been judicial reluctance to base a decision on inference when decision-makers could have been called as witnesses. There is, apart from exchanged correspondence, no contemporaneous documentation which I have seen which identifies any corporate thought process within Southwark about terminating the relationship or abandoning the software. The position is complicated additionally by the fact that there were many complaints originally pleaded but later abandoned and others not pleaded by Southwark. In the absence of evidence from the decision-makers, I would not readily have been in a position to find on a balance of probabilities that it was the potential matches complaint which was a sufficient cause of the decision to establish causation. Decision 115. There will be judgement for IBM and Southwark’s claim is dismissed.

Cited in 2 later judgments