Category Archives: research

Fiddling savings on shared services? Officialdom in need of reform

 By TonyCollins

An NAO report today suggests that some officials are fiddling projected savings figures from a shared services deal involving seven research councils.

It all began so well. A Fujitsu press release in 2008 said:

“UK Research Councils to implement shared services with Fujitsu. £40 million project will generate cost and efficiency savings across the organisations.”

An executive who representedFujitsu Services’ was quoted in the press release as saying at the time:

“Fujitsu is consistently proving that it can deliver effective shared services infrastructures and is playing a vital role in driving forward the transformational government agenda through shared services.

“Organisations that adopt a shared services approach can experience genuine economies of scale and reduction in costs which can be essential in their drive for continuous improvement.

Twenty-one months later Fujitsu and Research Councils UK parted company. The 10-year shared services contract began in August 2007. It was terminated by mutual consent in November 2009.

A revealing report, which is published today by the National Audit Office, shows how, despite the best intentions by the Cabinet Office to improve the management of IT-related projects and programmes, and decades of mistakes to learn from, some officials in departments are still making it up as they go along.

The worrying thing in the NAO report is not only what happened in the past – few will be surprised that the NAO report characterises the shared services deal as lacking professionalism. What’s worrying is officialdom’s more recent disregard for the truth when claiming savings for its shared services arrangements.

The NAO’s report”Shared Services in the Research Councils” suggests that officials manipulated – some could say fiddled – projected savings figures.

The NAO also found that officials awarded a £46m shared services contract to Fujitsu which came second in the bid evaluation. Exactly how the contract came to be awarded will be investigated soon by MPs on the Public Accounts Committee.

Origins of shared services contract  

In 2004 a review led by the Government adviser Peter Gershon suggested that the public sector should save money by sharing support services such as IT, HR and finance. In 2006 officials at the Department of Trade and Industry (now the Department for Business, Innovation and Skills) encouraged their colleagues at seven research councils to set up a shared service centre, which they did.

The UK Research Councils is an important organisation. In 2009/10 it spent £3.7bn, mostly on giving research grants to universities, the European Space Agency and other organisations. Its biggest recipient of grants is the Medical Research Council.

Fujitsu contract

Public servants appointed Fujitsu in August 2007 to put in place the ICT systems to underpin the shared service centre in a ten-year contract worth £46m. Fujitsu came second in the initial bid evaluations.

The NAO said that the bidding process produced a shortlist of three companies including Fujitsu. Said the NAO:

“The initial weightings applied by the [bid] panel had placed Fujitsu second: although the bid had scored well on quality, it was 19 per cent more expensive than the cheapest bid.”

An independent review commissioned by the project board backed the evaluations which put Fujitsu second. But the bid panel and the project board had concerns about the evaluation. The supplier chosen in the evaluation – which the NAO refuses to name – did not score well on quality requirements.

It appears that the bid panel and the project board preferred Fujitsu.

Mathematical error

Then officials happened to spot a mathematical error in the bid scoring. The corrected scoring left Fujitsu on top, as the new preferred bidder.

Said the NAO:

“… a mathematical error was identified by a member of the project team that changed the order of the preferred suppliers, leaving Fujitsu as the front runner

“The [bid] panel reconvened to discuss this but, rather than re-performing in full the quantitative and qualitative analysis and submitting this to independent review, it decided to appoint Fujitsu on the basis of a vote.

“In September 2007 the gateway review team concluded that the incident had weakened the value of the overall process and had left the project at risk of challenge.”

User requirements unclear

Full delivery was due in September 2008 but the project team and Fujitsu “quickly encountered difficulties, resulting in contract termination by mutual consent in November 2009”.

The NAO said there was “miscommunication between the parties about expectations and deliverables, primarily because design requirements had not been sufficiently defined before the contract started”.

Fujitsu consequently missed agreed milestones. “Fujitsu and the Centre told us that the fixed-rate contract awarded by the project proved to be unsuitable when the customers’ requirements were still unclear.”

Officials paid Fujitsu a total of £31.9 million, of which £546,000 related to termination costs. Despite the payments to Fujitsu, parts of the system were withdrawn and rebuilt in-house.

Overspend on Fujitsu contract

The NAO found there were “significant overspends on design and build activities and the contract with Fujitsu.”

At least £13m wasted on Fujitsu deal

Said the NAO:

“Had the Fujitsu contract worked as planned, we estimate that the additional £13.2m design and build costs … would not have been needed. In addition the project management overspend of £9.1m would have been lower, as, after termination of the Fujitsu contract, a significant overhead in managing contractors was incurred by the project.”

Fujitsu out – Oracle in

The breakdown in relations with Fujitsu led to the appointment of Oracle as supplier of the grants element of the project. “The contract with Oracle suggested that lessons had been learnt by the project following its experience with Fujitsu, with greater effort given to specifying the design upfront,” said the NAO.

Did officials know what they were doing?

In deciding how to share services the research councils came up with six options including setting up a centre run jointly by the councils or joining with another public sector agency such as one supplying the NHS.

But two of the options including the NHS one were dropped without proper analysis, said the NAO. The remaining four options were each given a score of one to three, against seven criteria. “The scores appear to be purely judgemental with no quantified analysis,” said the NAO.

Even if the six options had been properly appraised, the evaluation would have failed because it did not include a “do-minimum” option as recommended by HM Treasury.

“Overall, the quality of options appraisal was poor,” said the NAO.

Fiddling the figures?

 The NAO found that:

–         Initial estimates were of zero projected procurement savings from shared services. But by the time the first draft of the business case had been written the projected savings had soared to £693.9m.

–         When this project board queried this figure the research councils’ internal audit service scaled down the figure to £403.7m – but this included £159.3m of savings that internal audit had concluded were not soundly based.

–         Since the shared services centre began officials have recorded procurement savings of £35.2m against the business case and while of these are valid savings some are not. The NAO investigated 19 high-value savings that represented 40% of savings recorded to the end of 2010 and found that 35% “should not be claimed against the project investment”.

–         The research councils have been “unable to provide paperwork to substantiate the claimed saving”.

–         Savings claimed were indistinguishable from normal business practice such as disputing costs claimed by a supplier.

–         Clear evidence exists that the budget holder had no intention or need to pay the higher price against which the saving was calculated

–         Last month the research councils claimed that savings were £28m higher than they had reported previously owing to errors in the original numbers. But the NAO found that the councils were unable to reconcile fully the two sets of numbers; had not used a single method for calculating benefits or tracked these effectively; and had not included £7m of spending incurred by the councils. “Overall, this review has highlighted that Councils have not put in place proper processes to track benefits and forecast future operational savings,” said the NAO.

–         Further, investments needed to deliver projected savings have not been included in calculations.

–         Double counting. A revised target for projected procurement savings procurement “includes elements of double counting …”

Other NAO findings:

–        Four Gateway review reports of progress on setting up the shared services centre, including a review which put the project at “red – immediate action needed”, were not fully followed up. 

–         There was no evidence of intervention by the Department for Business, Innovation and Skills when it became clear the shared services project was likely to overspend.

–         The shared services centre has begun to match the pre-shared services payment performance of the research councils but a high number of invoices was on hold at the end of July 2011 because of problems with the end-to-end processes. About 5,900 invoices were on hold, awaiting payment, in July 2011, which was 21 per cent of all invoices due to be paid in that month. The reason for the delay was being investigated.

–         Despite the shared services arrangements, some research council staff were at times running parallel systems, or managing their businesses without adequate data.

–         In July 2011 the shared services centre had 53 key performance targets to meet but was only able to measure activity against 37 of them and of these met only 13..

–         Five of the seven research councils did not file annual accounts on time in 2011 in part because functions in the finance ICT system were not delivered by the project.

Some good news

Said the NAO:

“The grants function and its associated ICT system developed by the project has allowed the Councils to replace older systems that were increasingly at risk of failing. This is of critical importance, given that the processing of research grant applications lies at the heart of what the Councils do. The single grants system has the potential to make it easier for the Councils to collectively modify their processes in the future…”


The commendably thorough NAO investigation has shown once again how badly departments and their satellites are in need of independent Cabinet Office oversight when it comes to major IT-related projects. In that respect thank goodness for the Cabinet Office’s Major Projects Authority. But how much influence can it really have? How much influence is it having?

This NAO report suggests that some officials are fiddling the figures without a care for professional accounting practices. Double counting, not including full costs in projected savings calculations, not having paperwork to support figures and other such administrative misdemeanours indicates that some officials are making up savings figures as they go along.

What is to be done when some departments and their agencies are not to be trusted in managing major projects?

NAO report on shared services at seven research councils

Why this SME’s innovative ideas may help Nick Clegg understand the real causes of the riots

Could a new approach detect the early warning signs of radicalism in a way that ordinary research, surveys and intelligence gathering couldn’t? Or spot when programmes to reduce re-offending aren’t working?

In this guest blog, Andrew Moore, chief operating officer at DAV Management, whose customers include large public and private sector organisations, explains why government research into complex situations, such as the causes of the recent riots or making offender management more effective, requires a different approach that goes beyond supporting preconceived hypotheses to give new insights and fresh perspectives, and crucially, offers a means of detecting the early signals of situations that are developing in communities which can either be encouraged for the wider good or damped down before they can pose a threat.

Improving the Citizen Experience

An innovative approach to surveys and research

There is a currently a great deal of interest within government circles to determine how best to engage with its various stakeholders in different, more effective (and understandably less expensive) ways.

There’s the Big Society, the attempt to establish ‘happiness’ as a measure of the nation’s wellbeing (rather than just good old GDP), the need to engage citizens in a more direct and effective manner, with services designed around them rather than the structure of government.  In addition there are specific events that trigger the Government’s need to interact with sections of the population, such as the recent announcement by Nick Clegg that he wants to engage with communities affected by the summer riots in England, in order to understand who did what and why.  Then there are the government’s internal stakeholders – its employees, with whom it is seeking an altogether more symbiotic relationship – devolving power to the people on the front line who frequently know how to run services in a more effective and efficient manner.

Mutualisation’ and ‘Third Sector’ are terms that I suspect everyone is likely to become more familiar with over the next few years, even if now they may require some defining.  The long running debate about the future of the NHS is a very good case in point. And there’s the stated desire to get SMEs delivering innovation as part of effective government procurement.  As an interesting adjunct to this, let’s not forget that employees are also citizens, creating a fascinating cross-over of interlinked perspectives.

Of course there are other groups who may be thought of as stakeholders and these will have a very specific perspective on the delivery of public services.  I’m thinking here of offenders – those serving their debt to society and for whom the government is seeking ways to improve rehabilitation, reduce re-offending and become much more effective at identifying those most susceptible to radicalisation, extremism or self-harm. This of course has been brought into sharper focus by discussions over the severity of post-riot sentencing.

All in all, this represents a hefty agenda of public services reform and one which will test the government’s strategic planning and policy implementation ability to the max.  With such degrees of change being considered, it is encouraging to hear that government is embarking upon a listening exercise to garner the views of citizens, employees and service users, as some recipients of public services are now known.  Understanding what people want in order to deliver services they will use is a laudable objective, but what a task this must represent.  How on earth do you make this achievable?  Consider for a moment the potential population sample.  What constituency would you choose?  How do you get people to participate with sensible and meaningful responses?

Even if you can get all this feedback, how do you make sense of it?  How would you store, manage and interpret the sheer volume of data, relating to so many different aspects of life and stakeholder groups?  How could you be sure it doesn’t end up as an exercise designed to prove (or disprove) preconceived positions?  How would you spot the things that you don’t recognise – the identification of a strong belief system (that could make or break any changes in the way public services are delivered); the early signs of a rise in community ‘temperature’ that could lead to the kind of civil disorder witnessed in cities across England during August this year; or the indicators that some offenders are significantly more willing (and likely) to rehabilitate under certain conditions?  I could go on but I don’t want to labour the point.

It’s clear that in an exercise that will fundamentally change how most people interact with both central and local government, it makes sense to give those people a voice.  But this has to be in a controlled and manageable way, so that it is quick and easy to understand what that voice is telling you; gaining truly unique insights and fresh perspectives from which actionable decisions can be made and monitored that make a real difference to people’s lives, be they citizens, employees or service users – or, in some circumstances, perhaps a combination of all three!

Making people part of the process in this way is also an effective way of getting buy-in.  People are more likely to feel engaged, even if it’s by proxy (i.e. evidence of meaningful consultation establishes a degree of credibility) and by its very nature, changes the basis of the relationship between government and those stakeholders with whom it is seeking to engage.


So what’s the point in all this, why are these things being suggested as anything new?  After all, the idea of planning, shaping and delivering services against a well-defined need is surely common sense and is recognised as such by most people.  Well, as they say, the problem with common sense is that it’s not that common.  The truth is that the kind of knowledge and insight that is likely to be required by government in order to shape and deliver its vision for public services, is difficult (if not impossible) to gain from traditional methods and technology. A different approach is required.

What if you were able to poll large samples of the population on a variety of different topics and have the findings presented to you quickly and simply, in a way that wasn’t mediated by ‘experts’ and allowed you to interact directly with the data – at both a quantitative and qualitative level?  What if you were able to see things that you hadn’t expected; things that blew away commonly held perceptions about citizens or employees – giving you a clear and substantiated view of how people are feeling, what they really think of particular programmes and initiatives and how they are responding to specific policies and interventions?  Imagine being able to detect early opportunities to take action on a particular initiative that enabled you to maximise the benefits downstream or damp down a threat before it was even recognised as such.

It all sounds too good to be true, but advances in cognitive based solutions, using micro-narratives (snippets, stories, reports and other qualitative data) captured from samples of your target ‘audience’ and self-indexed by them to provide meaning from which incisive action can be taken, are turning these scenarios into reality.

The problem for strategists and policy and decision makers is that the environment in which they are operating is hugely complex; there are many small causes that interact and interweave to produce an end result, but no one cause is dominant.  The whole environment is continually adapting and changing and you can’t measure it at a point in time – it’s constantly evolving.  This is what’s known as a ‘complex adaptive system’.  It’s the kind of environment where outcomes are difficult to predict.  It’s highly sensitive to small changes, meaning emerges through interaction and, with the benefit of hindsight, you might be able to see where, when and why things have happened and how you could have dealt with a particular situation, but at the time it was erratic and novel.  Sadly, hindsight does not lead to foresight and processes to prevent a similar situation occurring next time will fail, because the next time things will happen differently.  The August riots in England were a perfect example of this scenario, where multiple small, erratic events interacted and evolved to produce a disproportionate, unpredictable and, in this case, catastrophic outcome, which the government is still trying to understand the cause of.

Complex situations frequently occur when you are dealing with people because they are inherently unpredictable and often driven by emotion.  The bad news for government is that, one way or another, people are at the heart of all of the major change initiatives and civil events that are currently under the policy spotlight.  You begin to get a sense for the scale of the challenge.  Not an overnight thing this.  [By the way, if you’re having difficulty getting to grips with the concept of a complex adaptive system think of mayonnaise.  If you’ve ever tried to make this from scratch you’ll know how uncertain it feels as the ingredients combine and the mayo gradually emerges.  One slip and it will curdle, the end result is never the same and it can’t be reverse engineered].

Fortunately, when trying to get to grips with a complex situation, a cognitive approach again comes to our rescue.  It enables us to probe the situation, sense what’s happening where and why and then respond accordingly.  It’s liberating for policy makers as it opens the door for innovation, enabling organisations to try things and see what works best in particular situations.  Fast feedback loops promote a low-risk, ‘safe-to-fail’ environment where those ideas that aren’t working are quickly identified and turned off, enabling us to get behind those that are delivering tangible results.  In this way, new services and new ways of working can evolve, meaning that the end result has a much higher chance of widespread adoption and, hence, long-term success.

The really good news for government is that game changing solutions of this type are really in the sweet spot when it comes to getting ‘more for less’, as today’s economy demands.  The levels of investment required are a fraction of the amounts that have typically been associated with major government change initiatives.  They are also much simpler to implement and run.  Once set up, data capture, analysis and reporting can be built into an organisation’s day to day operational processes, supporting (and stimulating) how it interacts with the citizens or service users it serves, or the employees it depends upon for the delivery of those services.  For example, making it part of how Offender Managers (previously known as Probation Officers) interact with offenders to try and reduce re-offending would be an excellent way to capture how the latter group is responding, say, to changing institutional attitudes and behaviours, revealing to what extent infantilisation (i.e. treating offenders in a condescending manner, as if still children) is being reduced.


A recent case in Canada illustrates how the solution can be implemented to improve the delivery of healthcare services.  In this instance, the authorities in British Columbia initiated a programme to help them understand the perspectives of all parties implicated in the unfortunate death of an elderly patient.  This had resulted from a breakdown in communication and subsequent decision making following the patient’s admittance to hospital suffering from congestive heart failure. Not unusually in these circumstances the single, sentinel event of the patient’s death was seen from very different perspectives by the various groups involved.  By adopting a cognitive based approach the authority was able to bring together front line and management staff to make sense of these conflicting perspectives and, as a result of the unique ‘safe-to-fail’ experimentation techniques supported by the approach, it was able to trial and subsequently implement changes in both policy and service delivery that will not only help to prevent similar incidents occurring in the future but also raised the quality of healthcare provided to patients more generally.

Just think how powerful such an approach would be for Nick Clegg in his quest to understand the complex human behaviours and emotions that came together to fuel the aforementioned riots in England.  And to have this at your disposal not only as a platform from which to take decisive action now but also to generate alerts when the ‘community temperature’ again begins to rise, must surely present a huge opportunity that any civil authority worth its salt would want to take advantage of. Instead it would appear that research initiatives are being launched by those with an interest in understanding and curing society’s ills that, albeit well-meaning and based on credible empirical evidence, may still ultimately turn out to be incomplete.  My concern would be that if a traditional research approach is adopted to try and make sense of what is essentially a complex situation (as I have defined above) then such initiatives risk revealing only those things that are readily recognisable and, having been mediated by ‘experts’, support preconceived hypotheses.  They are likely to  miss the opportunity to discern unexpected findings arising directly from the contribution of the people affected by (and involved in) the riots and fail to detect the early signals of situations that are developing in those communities, which can either be encouraged for the wider good or damped down before they can cause further unrest.  Addressing these issues by adopting a cognitive based approach will provide a much more effective feed into future policy decisions and social interventions.

If you’re new to the concept of cognitive based solutions it can be a bit of a challenge to get to grips with how they work and just what they give you, but once you’ve experienced the power of the knowledge and understanding that they deliver, you start to see applications everywhere you look.  The big advantage is that it’s easy to get started with low-cost, low-risk pilots that can start to make a difference to any organisation in a very short space of time.

To learn more, visit

Contact Andrew Moore at

Or call +44 (0)1189 974 0100