In January 2011 the UK Government set out to develop a new IT social security system in two years. Known as Universal Credit, this project sought to rationalize and consolidate the range of social security benefits that are available to British Citizens. According to the Department of Work and Pensions an agile development approach would make this enormous task achievable...and it would deliver on time.
You may ask how this could ever have worked, given that the public sector has such a distance to cover in terms of attaining agile practice. The Universal Credit scheme certainly looked like an inauspicious melange of hype and ludicrous ambition, and the dim bulb of Whitehall didn’t shed much light on practical details. However, in March that year, a new Government ICT Strategy was published which framed this initiative in terms of official policy. It was established that some form of agility was to lie at the beating heart of a switched-on and connected public sector. The Universal Credit project wasn't to be a speculative one-off. It was to be the start of an agile transformation across government services.
Even with this high-level commitment, it was difficult to view such heady aspirations with anything more than derision. The gap between policy and established government culture seemed just too great to make this transformation credible. It wasn’t long though before I started to have second thoughts about this. An agile government might indeed be unlikely…but surely it had to be possible. After all, I’d been involved in agile enterprise transformations that were hardly less challenging. I was buoyed, along with others, by the DWP’s observation that traditional suppliers weren’t up to the job, and that the field would have to be opened to SME’s. Shouldn't we therefore give this “Agile Government” initiative a chance? Admittedly it might just be wind, but at least the sails were filling. I posted a video outlining in simple terms how the venture could work. I observed that success was possible if there was flexibility of scope and if the contracts that were tendered reflected this. These contracts would have to be predicated not on fixed specification, but rather on the on-going delivery of value, while costs and time could remain fixed.
Unfortunately, no sooner had the ship left the dock than it started to founder. Despite its earlier rhetoric championing SME’s, the DWP reverted to type by awarding supplier contracts to the usual suspects including Accenture, HP, and IBM. Transparency around this selection process was very poor. The omens could hardly have been worse and for the next two years, the rumours of an agile shipwreck on the rocks of the DWP grew. Then, in May 2013, the Department announced that it was dropping “agile” from Universal Credit. This political distancing of the DWP from the strategy it had trumpeted showed that the initiative had indeed slipped beneath the waves. The question is why…and due to the continuing lack of transparency it is hard to come up concrete answers. What is clear is that of the 24 contractors engaged in Universal credit only 3 were SME’s. Beyond that we can only speculate, although any or all of the following are plausible.
- The “big hitters” who got the bulk of the work might have claimed that they could “do” agile, only to continue as before
- The contracts used to engage the big suppliers might have been improperly framed, and not fully conducive to agile practice
- Feature sets might have been essentially fixed, while the costs associated with Change Requests remained high
- SME’s with genuine agile experience might have been used as little more than window dressing
There is significant anecdotal evidence that most or all of these problems did apply here. If that is the case, then the deficiencies that stymied Universal Credit are not unusual. They are to be found wherever vested interests are threatened by an incipient agile transition, and particularly when the ancien regime decides to play for time. In my view, the next step must be to distinguish the formulation of agile contracts from the matter of actual procurement. Here’s why:
- Contracted terms do not prove that a signatory is competent to follow them
- Contracted terms do not prove the signatory’s intent
- Contracted terms might be ignored if a signatory does not fear repercussion
- Suppliers already working in an agile way can provide a far better indication of their abilities and intent than a contract ever will
- Procuring genuinely agile suppliers is likely to produce genuinely agile results…even with an imperfect contractual framework
It’s certainly desirable to get the right contractual terms in place, which is a matter I explored in an earlier post. My point is that even when this happens an agile transformation can still fail. Procurement of genuine agile suppliers is the real key. If you doubt this, imagine what might have been the outcome for Universal Credit, if only the DWP’s rhetoric about SME’s had been matched by honest action.