Public Sector Roundtable: To Share, or Not to Share?
This article originally appeared in Outsource Magazine Issue #25 Autumn 2011
As government-imposed cuts continue to impact upon the UK’s public sector, more and more voices are being raised in support of the shared services model – while an increasingly vociferous lobby is also mustering against it. To present a snapshot of some of the arguments for and against shared services, Outsource held a debate featuring three thought leaders with some very passionate views on the topic – and, at your service, we’re sharing it with you…
Outsource: OK Shimal: let’s get straight to the heart of the matter. Public sector shared services: good or bad?
Shimal Thakrar: In an age of austerity in which our public sector is in gear for major change after the moratorium in IT and Francis Maude’s review of IT spend in October 2010, IT suppliers have been summoned to look at ways to deliver technologies and information smarter. The public sector seeks a common mode of operation for access to data that is cross-fertilised across departments and organisations, available to those that request information. A shared services platform on which to enable access to such information therefore provides the key mechanics to ensure a merger of IT functions, thus creating economies of scale where the benefit of the savings is firmly put back into the hands of the taxpayer or spent on innovative ways to streamline further efficiencies in the age of information. The UK has the opportunity to showcase public sector delivery; now is the year to build that demonstration capability.
Nick Mellors: Optimising the use of staff, assets and other resources by sharing knowledge, skills and capacity is essential for the public sector to both live within its financial means and continue to deliver the services today’s citizens demand. Whilst turning this into practice has, in some cases, led to mechanistically imposed processes which fail to reflect the needs of the business (in corporate shared services) or the citizen (in shared delivery services); it doesn’t have to be this way. A mature approach to governance, the establishment of Service Levels which reflect the needs of the end users of the service and a professional focus on both the efficiency and effectiveness of the operation provide the key elements for a successful public sector shared service. Recognising the challenges of breaking down cultural barriers to change, political tensions and an increasingly sceptical workforce; the prize of reducing costs, increasing service levels and teams with increased professional pride make the development of public sector shared services both essential and good.
John Seddon: At the last Conservative Party Conference, I shared a platform with Bob Neill, Under Secretary of State, DCLG. Bob told the audience sharing services was a no-brainer. If, as is the case in Bob’s example, six Fire and Rescue Services have HR departments, then sharing them would produce savings. I asked Bob that if he is so sure sharing would produce savings, then why had the Department for Transport’s shared services programme gone from a planned saving of £54m to a cost of more than £175m – and the pain isn’t over yet. Of course Bob didn’t reply – but when you put that question to the protagonists they say “it’s OK if you do it right”.
But I remain interested in understanding the reasons for failure, lest we repeat the mistakes; and this becomes more urgent as we see an astonishing increase in examples of failures. I have learned to be sceptical about the promises of savings; all projected over many years, but no cash in the bank in the short term. And, as we have seen with the big failures, cash-in-the-bank doesn’t come through. It seems that the protagonists play to everyone’s no-brainer attitude with promises of jam tomorrow and by the time things have gone wrong, explanation rationalisations and blame replace cash as outcomes. Treating sharing services as a no-brainer turns the average brain off; evidence that ought to cause alarm just doesn’t compute. It’s time we got our brains engaged, especially Bob’s.
Outsource: John, what’s your response to Nick’s opinion that “public sector shared services [are] both essential and good”?
John Seddon: Nick, like many, argues sharing services is OK if you do it right. His argument for what constitutes ‘right’? Words about ‘governance’ and ‘professional focus’, which can mean whatever you want them to mean – and this whole debate is about method: what works? Which we get a clue for in Nick’s call for service-level agreements. But it indicates to me he doesn’t understand the problem. Service levels focus operational managers on meeting them, which is why we might pick up the phone in two rings but we don’t actually spend time solving customers’ problems. What I would describe as ‘classic’ shared services front and back offices work to these wrong-headed measures, meeting service levels but driving up the amount of failure demand (Seddon 2003) into the service.
Shimal argues that common databases will lead to better and more efficient public services. We have heard the same about the new Universal Credit (UC), from the DWP and HMRC. But how much would this cost and what value in real terms would such databases create? Before we plan to merge IT systems, without knowing the costs and consequences, we should consider other questions. In the case of UC, how many people on PAYE will be claiming UC? Does it make any sense to have everyone’s PAYE information on a database if only a few will be claiming UC? Moreover, where information about earnings and tax paid is relevant to benefit claims, what do we know about how well the current method for providing it works? In other words, what problem are we trying to solve – do we even have a problem?
I would draw Nick back to the question: what do you have to do to get it ‘right’? Challenging him on the role of service levels. And I’d ask Shimal to question: how do we know the provision of databases will the delivery of a service, other than as a guess?
Nick Mellors: I totally agree with John that setting the wrong service levels will always drive the wrong behaviours, similarly mismanaging any project will lead to failure. In the DfT shared service John quotes, the subsequent NAO report highlighted poor management of both the process design and the IT system development – although this might have been complicated by trying to bring more than one organisation together (hardly an unusual business occurrence) – these failings would have resulted in delays and cost overruns in any business change project, not just for shared services. Equally shared services can work in the public sector. In 2005 the Government Accounting Service Conference heard that ODPM had achieved annual savings of £2.3m, whilst reducing late payments by 14 per cent and supporting wider work that reduced the time taken to close the accounts from 41 weeks to 14 weeks – these were tangible achievements, not forward plans.
I remain interested in understanding the reasons for success rather than failure. Mature governance means each unit supported by the shared service understanding you don’t lose control of transactions or data simply because someone else is processing them and therefore you don’t need to keep detailed external checks and controls over the work of the shared service. I believe that many of the inappropriate service levels that John refers to are driven by immature governance models where business unit managers require detailed transactional targets rather than output or service targets.
Similarly where the shared services internal operational measures measure not only transactional volumes but also failure volumes (eg an analysis of calls to the service desk or returned invoices) improvements can be made to either the processes or the information provided to staff or customers to reduce these failures. This is the type of professional focus than can and does arise when a shared service management team manage both the efficiency and effectiveness of their business; but often doesn’t happen when small, typically under resourced, fragmented units of expertise don’t have sight of the whole operation and don’t have the opportunity to develop their professionalism.
John has said elsewhere “economy in service design is achieved through understanding flow, not scale”, I see no reason why we can’t achieve both. There are many reasons why organisations, public and private, are not as effective as they could be: the creation of a shared service is not, in itself, one of them.
Shimal Thakrar: I put it to John that whilst savings have not come to fruition the underlying requirements may not have been understood by the decision-makers. Was it that this modus operandi was sold to them as the next best delivery model, the future? “The competitor is building solutions, we must to remain competitive in the market” etc? Possibly. However, let’s analyse what we are suggesting here, a shared “service”. The decision-making should be made on what service is defined as, and what the customer/client expects from a service. I hope there is no debate that a service from a customer-expectation perspective is delivery on standards or some reference points against which performance is judged. Therefore, understanding what the customer expects is the number-one priority.
A shared service arrangement can go as far as creating some critical mass in grouping together the expectations of similar customers to create a price point that will be beneficial and of interest to making decisions in the client/supplier relationship. To decouple the two is what creates the increases in budget, project failures and excess spend and I believe this is what had happened in the case of the DfT: misalignment of definitions and expectations.
Bringing the definition and the execution together will result in a solution – a solution that can be applied, for example in the form of a database that is “fit for purpose” to those departments that feel they can access information from within it, feed into it and take MI from it with the confidence that it is consistent, accurate and above all relevant. Define the service, understand the definition and boundaries of the service, and we will have a public sector shared service platform that the world will look to review for their countries.
Outsource: Nick, do you feel that Shimal’s view on the customer service ideal has been achieved – or is achievable – in public sector shared services in the UK?
Nick Mellors: I think we all agree that understanding the service from a customer perspective is critical to developing successful shared services. I also believe that an emphasis on customer service is prevalent across the public sector with the vast majority of public servants actively seeking to deliver high-quality services the taxpayer can afford. (There is a very common public sector service paradox that says most people think public services are poor and inefficient – except when you ask them about services they have actually used, where there is significantly higher satisfaction.) However there are still too many examples of service failure with individuals repeatedly providing the same information to different parts of the same organisation or having to work across organisational silos to get the service they need.
For back office functions like finance/HR/ICT, my concern with the new ‘shared services agenda’ in government is that it doesn’t build on successes which typically came about through the pragmatic development of relationships, instead it talks about “establishing an equitable market of a small number (circa two) of accredited Independent Shared Service Centres (ISSCs)” built on the contradictory policies of “Independence is important to incentivise a better quality of services at a lower cost” and “Strong governance is essential and efficiency gains are proportional to the level of mandation in the use of shared services”.
Cambridgeshire and Northamptonshire councils recently announced their decision to voluntarily increase the scope of their Local Government Shared Service on the back of having banked £11m savings. If you have to mandate a service it isn’t doing what the customer needs.
John Seddon: When you dig down, Cambridgeshire and Northamptonshire’s savings amount to using less of their common resource – fewer properties and managers and lower IT spend. As usual, much larger savings are projected from lower transaction costs. These promises are not just illusory: they hide the truth that the real costs of service will rise.
Faith in lower transaction costs is at the heart of the shared services fad, which is derived from mass-production industrial thinking and designs based on shared call centres and back offices, connected by IT systems that control and report on the work. These designs focus on standardising work and managing activity (‘people’) in the belief that these will reduce costs. They don’t. When managers analyse how their systems really work, they discover that by failing to absorb variety, paradoxically their cost-driven industrial designs create huge failure demand and thus force costs up – which their day-to-day efforts to control make worse. The second revelation, however, is the massive scope for improvement. Having identified real customer demand, they can design services to serve it at the point of transaction: no back office required. No back office means no requirement for big workflow systems; no workflow system means no requirement for activity monitoring and so on.
I am about to go to Australia, where costly failures of public-sector shared services programmes are rarely out of the news. The attraction there, as here, was lower costs. When, if ever, ministers and managers become as obsessed with managing value they will see how designing for flow rather than scale drives costs out – always by much larger amounts than would have been considered achievable in any plan. By comparison, economising on common resources is insignificant to the point of irrelevance.
So to anyone considering sharing services, my advice is: study them where they are, then improve them where they are. (Do not alter the IT at this stage, although by all means turn it off.) These first two steps will deliver massive economies through flow and eliminating failure demand, in a matter of months. Then, thirdly, ask: can we make further economies by using less of a common resource? Finally: what, if anything, do we need IT to do to further improve the flow? The consequence is far less money spent on IT – and much greater value from it.
Outsource: So, to conclude: where do we go from here?
Shimal Thakrar: A lot of fair and thought provoking points have been raised throughout this debate. The public sector itself is a complex web that is governed not by directives but by prescriptive regulations that the Cabinet Office has set in an attempt to protect our citizens. However, as innovative technologies come to market the public sector must look to adopt a change in their modus operandi. Suppliers are keen to maintain their relationships with public sector organisations, and to demonstrate the options to allow data to be securely delivered to allow the sector to reduce their cost base for the privilege. A clear collaborate approach to delivering a shared service to achieve the benefits sought by both the customer and supplier is required: a showcase for the rest of the world. The time is now; the benefits outweigh the lengthy debates.
John Seddon: I recommend that anyone who is embarking on a conventional, IT-led, industrialised shared services project read Dangerous Enthusiasms by Gauld and Goldfinch (Otago Press). It is a thorough account of the failures of IT-led change. Twenty to thirty per cent of all projects fail completely; 30 to 60 per cent fail in part. And when you have read it ask yourself: do you think you will be lucky enough to be one of the few that succeeds?
Nick Mellors: I share John’s cynicism over the often-spurious claims in business cases for the introduction of SSCs where transaction volumes are multiplied by the difference between a ‘best in class’ benchmark and a current assessment of costs with little understanding of the relevance or composition of either (we’ll see how the Cabinet Office approaches this in October when it produces its Strategic Business Case for Central Government Shared Services). However, I don’t believe we need an ‘all or nothing’ approach to process improvement: the shared service model can and frequently does accommodate process variations where these are required to support the business – for example global finance shared service centres vary their processes to accommodate varying legislative requirements around the world. Whether you call them savings from using common resources or benefits from sharing services, there are positive gains to be made from the use of shared services; don’t let’s lose those benefits debating terminology or developing ‘pseudo-scientific’ measurement regimes.
About the Thought Leaders
Nick Mellors is Director of ISPB, a consultancy specialising in delivering public sector change and transformation projects. He has over 30 years’ public sector experience, and was a founder member of the central government shared services forum.
John Seddon has received numerous academic awards for his contribution to management science. He is the originator of the Vanguard Method, the means by which service organisations transform from conventional command-and-control designs to systems designs.
Shimal Thakrar has spent over ten years programme-managing Tier 1 IT change programmes that have involved elements of off- and nearshore blends. His career has evolved from PricewaterhouseCoopers BPO to Deutsche Telekom and of late as the Global Sourcing Engagement Manager for Atos.