HT-2020-000448 - [2024] EWHC 1185 (TCC)

DBS claims that it had to retain a number of temporary staff members to perform Barring work between October 2016 and September 2017 due to the delayed delivery of R1 Barring. This is Schedule of Loss item 3. There are also loss of anticipated savings claims relating to R1 Barring Portal Quality Issues (Schedule of Loss item 11) and to R1 Barring Standards/Quality Costs (item 24).

The claimed sum is £1,971,434. Mr Hain’s assessment is £2,424,934.

As explained by Mr Hain in his First Report, DBS’s claims for loss of anticipated savings in respect of R1 Barring rely on two documents: the Workforce Planning Model (or ‘WFPM’) v15.6 and an analysis of staff costs incurred (‘Barring Payroll Analysis’). DBS contends that these two documents provide the Court with (1) DBS’s best estimate of the efficiency savings expected had a compliant Barring Portal and Solution been delivered (the WFPM) and (2) the difference between workload prior to and after delivery of the Barring Portal and Solution.

Mr Hain also explained that DBS had estimated the lost efficiency by carrying out the following assessment:

As such, the FTE %Removed and FTE %Decreased calculated by reference to the WFPM is a fundamental part of the assessment.

In relation to the delay period, the loss is explained as having been calculated as the anticipated savings for the six month period from October 2016 to March 2017, plus 50% of the anticipated savings for April 2017 to March 2018, approximating to the period from April 2017 to September 2017.

In his witness evidence, Mr Gergely, who was a central witness in relation to the WFPM, said as follows (para 34):

Whilst accepting it was not ‘a perfect reflection’, Mr Gergely considered that it was nevertheless usable, in his view, for the purposes of the loss of anticipated savings assessment.

No doubt at least in part on the basis of this evidence, there was, as DBS points out in its Written Closing Submissions a measure of agreement prior to Mr Gergely giving evidence at trial as to the expected efficiency savings on the basis of the WFPM. As recorded at Section 6.3.1 of the Joint Statement, Mrs Wall recorded that,

Mr Gergely was however crossed-examined at some length on the chronological development of the WFPM. The starting point is that in around August 2015 (i.e. before any UAT), DBS performed an initial assessment of the changes which would occur in ways of working within DBS following R1. This was known as the ‘Detailed Level Organisation Design’ (or ‘DLOD’). In around early January 2017, DBS used the WPFM to update the DLOD following R1 Barring Go-Live. A column was added to identify whether a task was expected to be removed, be retained, decrease in effort or increase.

As TCS submits, and I find as a fact, it was known at the time that this was an inadequate exercise. The WPFM included important caveats, all of which were agreed by Mr Gergely:

In April 2017, someone called Marta Bergonzini from DBS produced a presentation which stated:

It repeated on slide 5 a number of the caveats and on slide 7 set out planned improvements to the ‘rough view’:

Slide 8 indicated that there had been little validation of duration of activities and ‘the level of confidence is low’. It is particularly relevant that DBS were themselves indicating that using actual data collected from the first few months of operations would improve future headcount estimates. Mr Gergely agreed in cross-examination that it was clear, at this point in time, that a new version of the WFPM needed to be created.

DBS’s Operational Organisation Design Document v0.2 dated 14 June 2017 indicated that it had not been possible to assess the full end to end functionality of R1 for the purposes of workforce modelling, and that DBS anticipated that as it moved into the next phase of R1 delivery (training for staff), it would be possible to better assess the impact on workforce and the ability to realise the theoretical business benefit. It was recorded at this time that this would form the basis of the next review of this document for Operations (Barring).

The DLOD dated 13 July 2017, indicated that, ‘a number of additional processes (or temporary workarounds) will need to be added post R1, until all operational areas have gone live. The exact impact of the decreases (in terms of potential FTE reduction) require further investigation by Business Analysts with the required level of R1 knowledge to determine any time savings. As yet it has not been possible to make any accurate assessments within the testing environment.’

There remained, therefore, a significant amount of uncertainty about the estimates, as explained during cross-examination by Mr Gergely (Day 13/49):

In October 2017, an internal DBS presentation indicated the work that it was then anticipated was required in order to revise and correct the DLOD assumptions, as reflected in the following slide:

Mr Gergely accepted that these tasks would need to be carried out in order to work out what the anticipated savings might be (Day 13/64):

In relation to a review Mr Gergely undertook in October 2017, he also gave the following evidence about the (un)reliability of the document as then perceived contemporaneously (Day 13/55):

Mr Gergely undertook a review of the WFPM following Go-Live. He was cross-examined in relation to a document produced after Go-Live in which an number of the assumptions in the 19/01/17 WFPM had not been borne out. One of these related to the Disclosure Information Team (or ‘DIT’) in respect of which the document showed an anticipated removal which had yet to occur or, indeed, in respect of which activity effort had increased.

This line of cross-examination led to the following exchange in which Mr Gergely explained that, with the benefit of hindsight, the document did not produce a figure which DBS could rely upon (Day 13/60):

This fair evidence merely reflects, in my judgment, the view expressed by Mr Gergely’s email of 15 November 2017, following his contemporaneous review, when he wrote:

Equally important was Mr Gergely’s evidence confirmed that it was after DBS had a live system that they could properly assess the work required to produce Barring Case work (Day 13/65):

I accept the submission of TCS that, on the basis of this factual evidence, Mr Gergely effectively disavowed the evidence in his witness statement that the WFPM was a ‘usable estimate’, at least in the context of the claim DBS is advancing. On any view, in my judgment, it is clear that, even as a working or rough estimate at the time, it suffered from considerable uncertainties and was shown, shortly after Go-Live, to have contained flaws, none of which have been addressed or corrected by DBS in its use of the document to underpin its claim. It is right that Mr Gergely had already accepted at paragraph 36 of his witness evidence that post Go-Live, he was concerned that the WFPM was not a suitable baseline to calculate the R1 impact because the processes were so different: however, this does not mean that the Court should conclude, as DBS contend, that the WFPM nevertheless is an appropriate basis upon which to calculate a delta which reliably represents a loss of anticipated savings. In my view, it plainly is not.

Whilst the foregoing conclusions apply to all the loss of anticipated savings claims, the particular context of Updated Schedule of Loss item 3 is the delay claim. In the context of a delay claim, the use of an estimate of savings is particularly problematic, even if it were reliable (which in this case, it was not). This is because, as Mr Gergely accepted, there would exist better evidence as to the actual time taken by tasks post-Go Live which can be compared with actual data about the amount of time taken in relation to the R0 system. The starting point for the quantification of the actual loss caused by the delayed introduction of savings can (and should) be calculated by reference to the actual savings that were, because of the delay, not (actually) generated as soon as they would have been. Taking a simple example, if post Go-Live the evidence demonstrated the efficiencies from the introduction of R1 actually generated an ability to reduce the workforce by 10 FTEs, then this is the lost actual saving caused by the delay for the relevant period of delay (no doubt taking into account the fact that the actual savings may not have been immediate). Even if the WFPM was not of itself unreliable as an estimate (as it plainly was), a claim for actual losses caused by a delay should be assessed by reference to actual, not estimated, losses when the evidence plainly exists to have permitted such an assessment.

Moreover, it is plain that even if (contrary to my conclusion) it would have generally been appropriate to advance an analysis based upon the (unreliable) WFPM in order to generate a percentage expectation of reduction which is then applied to actual costs, the evidence of what actually happened during the period of the claimed delay demonstrates that it is clearly the wrong approach on the facts.

Mr Gergely confirmed in his witness evidence (at paragraph 35) that DBS had already allowed staff numbers to drop by around 48 people in anticipation of making savings, as shown in the WFPM v15.6 as at 31 October 2016 (i.e. the commencement of the period of delay). Mr Gergely also confirmed in cross-examination that the figures of actuals within the WFPM were accurate. Although, as he continued to explain in his statement, DBS considered taking on short term agency staff in the lead up to Go-Live in order to make up for the reduced headcount, DBS decided to stick with the numbers they had, and to rely upon DBS’s fixed-term appointments and staff overtime in order to meet the workload. There is no evidence, therefore, that any people were hired due to the delays, in circumstances where DBS had already put into effect what it anticipated to be the staff savings. Mr Gergely’s factual evidence was that 12 people were then added after Go-Live rather than before.

There is no basis, given these facts, to attempt to assess delay prior to Go-Live by anticipated savings by reference to the theoretical percentages applied to actual costs derived from the WFPM. The evidence is that anticipated staff savings had already been made. One basis of claim could have been, on the basis of this factual evidence, based upon the overtime paid to the (reduced) workforce because efficiencies did not commence as soon as anticipated. That was, after all, the manner in which DBS actually sought to address the problem caused by delayed introduction of the new systems. But no such claim has been advanced and the Court has no proper basis upon which it could carry such an exercise out.

In the circumstances, because of, but also irrespective of, the unreliability of the WFPM in terms of its forecasts, I conclude that the claim for anticipated savings as a result of R1-B&B as advanced by DBS is fundamentally flawed and fails.

DBS’s claim for Loss of Basics Contribution has not been pursued.