The status quo has arisen from a broad assumption that if policy and guidelines are available they will be found, read, understood and followed. The evidence suggests that in times of high operational delivery pressure, combined with a shrinking public sector workforce, this is not the case. Previous efforts to change behaviors have involved little more than ‘restating the rules’. Where attempts have been made to introduce technology to address the problem it has often been counter-productive; implementing old-fashioned Electronic Document and Record Management Systems with little regard for how users work - demanding significant extra effort from the workforce to manually create extensive metadata and\ make decisions about what to keep, where to keep it and for how long.

This innovation is different because it is driven by the perspective that a benefits focussed value proposition is the most effective way to support behavior change, coupled with harnessing emerging thinking in behavioral economics and behavioral science to ‘nudge’ civil servants in the UK to improve information management behaviors. Where technology and techniques already exist, such as the use of e-discovery in the legal profession, the innovation comes from discovering how to apply it on a broader and more diverse problem space across government, at significant scale and as part of ‘business as usual’ processes.

The technological innovation in exploring the application of text analytics, natural language processing and sentiment analysis in real-time to support information management and exploitation is, to the best of our knowledge, a step-change in the public sector consideration of such technology (with the exception of some defense and security contexts). The concept of combining this with the development of a cross-government semantic knowledge graph is, again to the best of our knowledge, new to the public sector. Inspiration has not been drawn from any specific other innovation efforts; it has come from a long term tracking and analysis of technological and societal trends and creative consideration of how they might inform solutions on this problem space.

The innovation consists of a number of projects arranged in different phases designed to be mutually supporting in delivering the overall program objectives and benefits. Different elements of the program are therefore at different points in the innovation lifecycle. The problem was identified in the course of an independent review into government record keeping and information management. A ‘discovery’ phase was employed; clarifying the problems, understanding the needs of government departments and identifying areas of investigation for future solutions. An ‘alpha’ phase followed, looking at innovative approaches to digital legacy management by piloting the use of e-discovery and data analytics software, then sharing lessons learned among departments.

The need to tackle long-term compliance improvement and behavioral change was supported by developing a ‘Knowledge and Information Behavioural Change Toolkit’. Exploratory workshops have been used to reframe the program to strengthen the current value proposition and increase the scope of technological innovation. Success metrics are being developed with the concept of a whole of government risk dashboard being the significant quantitative measure. The extent of behavior change achieved and shifts in staff satisfaction with their ‘digital workplace’ will be assessed via longitudinal qualitative surveys, The spread of the innovation is occurring through cross-government networks such as the Knowledge & Information Management profession. This includes helping government departments learn from both the central team and each other in a collaborative setting.

Collaboration with The National Archives has been essential to bring deep expertise in archival practice into the team, especially in the digital domain where they have a world-leading digital archives approach. Similarly, the contribution from the Government Digital Service has allowed the team to engage in the implementation planning for major information systems across government, providing for the opportunity to support good information management practices from the start. Technical contributions have also come from the Government Legal Service and the Crown Commercial Service. Broader collaboration with the cross-government Knowledge and Information Management profession has been essential as responsibility for addressing this problem space is distributed across government departments.

During the discovery phase workshops were undertaken with consulting firms, highly-regulated industries, and academics to understand best practice in information and knowledge management. The team has engaged with small-scale providers of data analytics solutions and larger system integrators to identify available ‘off the shelf’ solutions and understand how these might be deployed at scale. Commercial stakeholders were invited to formal ‘Supplier Engagement Day’ and a sequence of 1 to 1 discussion with technical and procurement specialists. Civil society organizations and citizens have been engaged via a mix of formal briefings (e.g. at the Information Management and Records Society annual conference) and informal activities such as discussions at the community of practice meetings of civic-minded technologists. The academic community, a long-term consumer of government information through archival research, has also been engaged through conferences (e.g. at Northumbria University).

Early results have confirmed that the use of e-discovery and data analytics can significantly reduce the volume of information that needs to be reviewed by more manual approaches. Though the exact process for how this will work at scale is still in development, the use of data analytics is expected to make it more likely that government will be able to manage its legacy digital information from a risk perspective, as well as gaining valuable insights from it. The behavior change toolkit, developed in the alpha phase, has already been credited with significantly improving the information change strategy of a major government department. If some of the more stretching goals of this work (record categorization and network creation) come to fruition, the upsides for government productivity and outcomes are potentially very significant. Though it is too early to speculate on the exact scale of impacts and returns, we would hope that civil service working practices would change considerably for the better as a result of our work, with potential knock-on effects for outcomes for citizens and customers of government services.

To date, we have encountered a mix of procurement, security, organizational and cultural challenges. Some of these have resulted from commercial difficulties arising from interaction with major IT system integration and operating companies. Others have included the difficulty of engaging senior leadership attention on this problem for a period of sustained effort. The team is exploring a variety of ways to address these challenges, for example by exploring ways to bring risk which implies future effort (and therefore resource) forward as a nominal contingent liability, and (in a different context) re-framing the problem to focus more on harnessing value than managing risk.

Pre-requisites for success in this type of innovation include sustained senior leadership engagement; aligned motivation and values across the team; and, the ability to apply different levels of pace and delivery expectation to different components of the solution set depending on complexity and the urgency of the corresponding problem area.

The potential for this innovation to be replicated elsewhere is very high. The problems are generally organization agnostic and exist in any large (>500 knowledge workers) organizations - especially if they are federated to any extent. Thus similar problems are found in local government authorities and health service organizations. We hope that we will be able to share lessons with other governments, charities and other not-for-profit third sector organizations. The technology approaches are being pursued in a deliberately vendor-agnostic manner. As government departments use a variety of cloud-based office productivity systems it is essential that the tools and techniques we develop work across different systems.

Dealing with complexity in a problem space of this range and scale works best with an open and collaborative approach - but this is only sustainable if the senior management of the contributing organizations can align themselves with a common set of goals which also support their organizational objectives. Success and progress also require sustained effort, ideally from a central team which can coordinate and drive activity and ensures alignment and effort from all partners. A hallmark of this work has been experimentation and testing. It can be difficult to persuade stakeholders to experiment when resources are limited, but this project has had success through an incremental approach.