All organisations, whether in the private and public sector, have been under constant pressure to save money. In the past, this often led to outsourcing business services to a third party, seeking to adopt a shared service model, or embarking on an ambitious IT transformation project that is meant to solve all the organisation’s problems. Unfortunately, while project failures are not unique to the public sector, it had gained particular notoriety for such high-profile failures, particularly when it comes to IT implementations. In addition, public sector procurement has been often overly focused on cost reduction, without adequate focus on the long term value potential to improve agility, resiliency and flexibility too. The introduction of G-Cloud as a new procurement framework and GDS to help drive the digitisation of public sector services with CTS to ensure effective use of IT, was meant to bring an end to the era of large monolithic contracts, of reliance of outsourced skills and technology provision, and of projects that failed to be on cost, to be on time or to meet requirements. In addition to the contractual challenges that he faced on arrival at DVLA, Patterson faced similar issues when he was at CTS, working with departments to provide the ‘right’ technology for Civil Servants across government. In most cases, he found that the technology required could be built, but more significantly purchased via G-Cloud in the market once, and then used many times across multiple departments. However, he found that the main barrier preventing departments from investing in these solutions was still primarily the contractual landscape for the system maintenance that had been in place for many years. Many departments still had large, legacy contracts using large system integrators which affected their ability to change their technology estate. Contracts that treated IT as a ‘static’ item when in truth the technology was constantly moving forward and improving. Departments were faced with costly change control requests and complicated workarounds to link up cloud-based commodity solutions with their existing technology. We asked Patterson if he thought that the public sector had learned from the mistakes of the past:
Some of the work in CTS, for the first time in UK government was providing new data, and transparency. We began to map every technology contract, by department, and by supplier and also mapped the supporting supply chain, the internal commercial capability and, for the first time, create a full technology understanding of what currently existed pan government, by department, by product, version and level of customisation. There are interdependencies between capability that the organisation has in commercial, technology and service design. In conjunction with the current commercial landscape, how you commercialise it, your current architecture and your change portfolio. Historically contracts or outsourcing IT assumed that all these items were static, but a process would be commercially drawn up that would allow for change to happen, so as soon as the contract was signed it would be subject to constant change control even before the initial ink was dry. The reality is that all these plans are truly dynamic and that this needs to be considered when you want to transform services, technology and people. What was needed was a new model of contract, one that baked in the Service Standards, and Technology Standards, considering that these would continuously need updating. Key to this would be that SI’s would know, from the get go, that the requirement of any commercial agreement was that technology is fluid, but regardless of this, the goal would be for products to adhere to standards rather than just new technology items ‘new shiny things’. It was always disingenuous that System Integrators were ‘bad’. Logically, government was forcing SI’s and departments to sign outdated types of commercial agreements that would not exist in the more modern parts of the private sector. Much of the focus was in ensuring the process of getting a contract in place and on ‘outcomes’ as if they were commodity purchases. Instead, the focus should be on the necessity for an agile and dynamic capability, where the builds and architecture allow for disaggregation and the creation of microservices, by ensuring that any development allowed for API connectivity and reuse. Patterson believes that there is a lot that can be learned from the private sector:
Transformation is often a misused term -in truth, very little is transformational, and we actually live in a world of continuous change and change opens the door to disruptors. Since 2005, we have really seen digital as a major disruptor where greenfield – new companies that have been able to scale rapidly using digital concepts – have thrived because they’ve avoided the corporate and technology legacy their traditional competitors have built their businesses around for years. Successful new companies come from a user centric perspective in order to grow very fast, mapping and using user needs to deliver services which they know users want, and are designed and built accordingly, and to do so in an elastic manner using technology and data to drive new propositions that they know users will want. When you translate that into the public sector, what you realistically have is the perception of a captive market that is still dictated to by policy and process rather than being user centric. With the birth of GDS, a strong leadership in position, and a climate of austerity, it was a perfect storm due to the need to be transformational. The main issue is that the Public Sector currently has an outdated mode of business case creation that historically lent itself to a large big bang approach and a waterfall delivery methodology, when in reality, both the Treasury and Departmental Finance Directorates need to adapt their rules and criteria to enable the departments to accelerate their digital transformation whilst ensuring that there was the necessary financial rigour. Controls were needed that are fit for purpose in the Cloud and the Digital era, and for the rapid change that is now required to advance policy change in government. Unfortunately, the Savings Review structure seeks savings of operation budget in the form of revenue spend, but the move to Software as a Service (SaaS) and Cloud Services means its capital that is saved, not revenue!
Patterson recognises that real progress was made under the initial GDS regime, not only with the high-profile exemplars, but in pockets across government. However, whilst he is a strong supporter of the technical standards and believes that disruption of the technology employed by government was necessary, he regrets that a single new architecture for government was not defined at the outset and that central control has since been lost. Patterson feels that the Digital agenda, which was moving at pace under Bracken, was impeded by the lack of a clear Technology Strategy being in place to support it. He believes that Tech should have done far more than just create standards.
Technology was the highway on which Digital Services can be driven, and the Technology function didn’t do what it really should have done to accelerate things, and design to unify departments. Patterson remains a strong advocate of the Controls process, but believes that it requires a coherent and strong centralised technology architecture direction to support it. The problem is that digital transformation is not a one-off project, it needs to be an ongoing process. For it to be effective, it cannot be done by individual departments in isolation. True service reform also needs to be both cross government and citizen-centric.
Patterson believes that part of the problem also stems from policy creation:
There still isn’t enough alignment between policy and delivery. All too often, policy is developed with little thought to its implementation. Universal Credit was a great example of this. Policy needs to be applicable to the same disciplines as technical delivery, user centric, and designed for simplicity and efficiency. There are exceptions to this such as the work Kit Collingwood- Richardson did in placing users at the heart of policy making. Even so, this seems still to be unique, and not either common place or mandated. Mixing Users, Design, Policy Making and Tech together will always yield better outcomes.
Without this holistic approach to policy making, there is often a conflict of policies. One policy may be focused on reducing cost by commoditising and optimising a service, while another policy may focus more on improving a service by tailoring it to the requirements of a particular department or personalising it to make sure that it is more user centric or even seeking to integrate it with their other services, all of which would conflict with efforts aimed at commoditisation and optimisation.
Patterson recognises that progress has been made in terms of getting away from the large, inflexible outsourcing contracts of the past:
We have made great strides in moving away from ‘Big IT’ as it’s called, away from monolithic architectures, but there is still a massive overhang from the large outsourcing contacts of the past – many of which still have some time to run, or have even been renewed because civil servants are too busy with Brexit to focus on new and better value options. This can mean that we are also still bound by inflexible contracts with external suppliers: The interrelated nature of transformational changes cannot always be completely foreseen at the beginning. Contracts and supplier relationships need to be flexible and open. Inevitably, there will be issues that were not anticipated at the start and cooperation and flexibility is crucial if you are to overcome the hurdles and challenges.
Also, government has not yet learnt how to actually utilise the System Integrator and Vendor Market to its advantage. It needs to create a viable ecosystem of SI’s, SME’s and skilled Civil Servants that rival the Private Sector in skills as well as experience. As long as government separates the Commercial, Capability and Technical aspects by ‘departmentalising’ the disciplines, it will continue to fail to successfully transform sufficiently to become truly digital and serve users in the way a modern government will need to. The day to day expectations that citizens have in terms of customer service from their digital interactions now far surpass those that they receive in the Public Sector. GDS kick-started the progress towards meeting some of these expectations, but the Civil Service has so much more to do. Citizens do not differentiate between departments and service silos as the Public Sector does. They don’t care whether it is Central Government, Local Government, or District Councils. Citizens just see the services as singular items, not federated. Imagine buying your goods on Amazon, then logging off and signing in for the delivery, it would make no sense. Services need to be designed end-to-end from the user perspective, instead of being designed to meet the government or local authority’s historic, legacy process. Part of the outsourcing overhang relates to skills:
Skills and knowledge transfer remain a problem: Basically IT was seen as ‘not our core business’ and, like many organisations at the time, it saw outsourcing as the way forward. The fundamental difference between what the Public Sector did in comparison to the Private Sector was that the Public Sector used outsourcing to bypass the salary and training costs for IT professionals, and the private sector retained control over IT by retaining well paid technology leadership to oversee its architectural direction. Government has historically failed to invest in the necessary skills, then in squeezing the Sis, it forced them in turn to cut back on training for their people as they sought to maximise efficiency and minimise cost. The previous reliance on outsourcing hollowed out the technological skills base within government, with departments that turned to service providers to overcome skills challenges finding that the service providers were recruiting from a limited pool of skills and often poaching from these same departments in order to provide services, thereby further exacerbating the skills shortages within the departments that they served.
Knowledge transfer to properly skilled resource is vital for a successful transition of any service. In addition to technical knowledge, there is information about everything from processes, procedures, tools and governance as well as infrastructure and security, and even this doesn’t recognise the investment in culture and relationships. Managing the incumbent provider is not easy – the incumbent may not want to provide the information and knowledge, making the exit strategy hard to manage. The new service provider will also be unable to provide accurate SLAs
and other performance measures if it doesn’t have complete and coherent understanding of the interrelated processes and systems.
In summary, Patterson explains that:
The UK public sector is still reliant on third parties for skills, having never really invested in its people. The Civil Service sees IT as a generic skill set and as long as it continues to outsource, it will continue to get its tech skills via contracts and contingent labour, rather than via headcount – never really either a long term-oriented or cost-effective approach.
On account of the ongoing frustration at a senior level, we have seen a string of resignations as many of the most experienced technologists, and those with digital expertise, decide to quit the public sector. The decentralisation and weakening of the GDS agenda have also given permission for Departments to return back to a siloed mentality, which is the worst of scenarios given that Brexit is going to require a centrally driven approach, with strong experienced leadership in the Digital, Data, Delivery and Tech space.
All in all, Patterson believes that a great deal has been achieved. He worries however that the relaxation of central control (and to some extent the distraction of Brexit) along with the loss of critical skills may see some return to blind-outsourcing rather than what he calls ‘smart sourcing’ – something that he cautions departments to be aware of if at all possible.