I will begin with a quick critique of some of the original story. The Cabinet Office did not have an elite team of IT experts to loan to DWP to help sort out the Universal Credit. The Major Projects Authority within Cabinet Office is a team of reviewers, not a team of trouble shooters. The Government Data Service has only just started to recruit those it needs to expand its project monitoring capacity. Even when they are in post it will have too few of those with experience of planning and delivering major projects to provide serious support for the change programmes of the great "Silos of State", as opposed to the "dashboard" monitoring of those programmes and assistance for smaller departments with less ambitious projects.
Meanwhile Gov.uk is still best viewed as a set of design principles and some "lipstick on the face of a pig", although hopefully it will contionue to make serious progress over the year ahead as it gathers momemtum. In parallel the GDS is trying to build the necessary in-house expertise in open source and cloud services to help make a reality of its aspirations for a brave new world of re-usable mix and match modules for common needs, using rationalised, resilient and secure data centres.
Meanwhile the DWP needs very different skills to handle a massive change programme where the main investment is organisational change and staff training, including to handle co-operation and secure, accurate and reliable information exchange across silo boundaries (e.g. with the Border Agency, Local Government and HMRC). By comparison the techology requirements and even the data volumes are relatively trivial, compared to those at Vodafone, from which DWP has recruited its new of Head of Digital (who was undoubtedly dismayed by the lack of hand-over when had arrived). The problem lies in the complexity of the assessment of the individual cases by front line staff, before the data is put into the main delivery systems.
Unfortunately, over the past two decades politicians, officials and public sector consultants have persisted in putting digital carts before "people process" horses. This time, after the Minister put his foot down and called a halt, we saw squeals from technophiles, in cacophony with squawks from the Public Accounts Committee, neither trying to identify why and how the programme had got out of control, despite supposedly clear agreement at the start that the implementation of Universal Credit would follow a low risk, incremental strategy, based on identifying and testing claimant "pathways "
The best news is that Ian Duncan Smith has not called in yet more consultants. Instead the Department is trying to rebuild the in-house skills base to do that which cannot be sensibly contracted out, whether to an outsource provider or to Cabinet Office (even if it had the necessary skills available).
The successful use of agile methodology requires that the users are intimately involved, ideally trained to use the tools themselves, with the IT "experts" in support roles only, until such time as systems are ready to be scaled for live running. Then the IT experts often have a major role, "fine-tuning" the systems (o ensure rapid and reliable response for the volumes to be handled. In the past that fine tuming has often required a complete re-write, while keeping the user interface and application definitions the same. That is less necessary today, using tools developed originally for automatically transitioning legacy systems to new operatiing systems.
Over the past two years we have heard much rhetoric about rebuilding the skills of Government to be an intelligent, or at least less stupid, customer. Many of us feared it would prove to be hot air, botched and/or rushed - e.g. with inexperienced staff trying to run before they could walk. I am hopeful that I will soon be proved wrong. Both Ministers and officials appear to have come to appreciate that the efficient and flexible delivery of on-line services to meet changing needs requires giving power to front-line users to work direct with in-house IT support staff, not binding them with inefficient and inflexible outsourcing contracts and overseas call centres, whatever the nominal savings from centralisation and standardisation. That is also by far the best way of developing the skills on both sides. It is also usually easier to give the necessary IT skills for most of the roles in incremental project planning, development and implementation s to users than to teach IT staff how the organisation operates.It is the skills to organise the process that are in short supply and the Cabinet Offiep lans for work in this area are most welcome.
It is interesting to see how some of the former outsourcing dinosaurs have seen where this leads. They are beginning to look at how to provide scalable item of service, open source and cloud services to the SMEs and to the Third Sector providers to which HMG says it wants to devolve delivery to those who are hardest to reach.
Unfortunately, the Government Procurement Service still appears wedded to the idea that centralised procurement (as practiced by supermarket and clothing chains when screwing British farmers and exploiting Bangladeshi sweat shops) will give value for money. And we should not under-estimate the challenges of changing that approach, given Treasury accounting rules as well as EU Procurement law. In consequence those with innovative, joined up solutions, that really would deliver better service at lower cost, have simply walked away - to sell to those who appreciate what they have to offer.
The rebuilding of Central Government's skills to be an intelligent customer may have begun it is a long long road.
We can, however, see the likely future.
There will be a Government Data Service (akin to the old CCTA) that
- helps smaller departments with their technology needs (including the use of open source, cloud based systems to meet common needs)
- sets standards for inter-government co-operation (including identity arbitrage and secure data exchange)
- monitors progress (in co-operation with the National Audit Office).
In parallel the great silos of state, who collect and spend most of the money (HMRC, DWP and MoD) will rebuild their in-house expertise and run down their all-encompassing outsource contracts, replacing them with multiple sourced, item of service frameworks.
In a post Snowden world we can also expect to see increasing pressure for any outsourced UK public sector services to be supplied from locations or suppliers which do not fall under foreign jurisdictions which claim right of access to data or communications and/or meet all UK regulatory, security and resilience requirements. Given similar pressuress in other EU member states this may also mean that the Digital Single Market, particularly for public services, will remain a fiction for the foreseeable future.
How long the process of change will take is much less easy to predict.
How many of the current dominant suppliers will fight to preserve the status quo and/or milk their legacy contracts for as long as possible? Perhaps in cahoots with a Government Procurement Service for whom such a change of approach represents an even greater re-education challenge?
How many will join those who have already walked away from UK central government as unprofitable (after sales costs, overheads and hassle) despite being over-priced?
How many will they join those looking at how to do business in the new world that is emerging?
Some are, of course, trying to have their cake and eat it. They have set up innovation teams looking at new ways of doing profitable business (e.g. providing item of service support for flexible and selective outsourcing via a variety of channels) at arms length from the teams seeking to extend the life of their current "big" outsource and PFI contracts and win new ones.
Like so many other strategies, this is far easier said than done. I would, however, far rather have shares in those who follow such a twin track approach, are open about doing so, and allow their teams to compete head to head - including for the in-house resources to deliver that which they win. That is in no small part because I suspect that those who pick too soon will go down just as painfully, albeit perhaps faster, as those who move too late and that getting the timing right will be impossible.