Government departments and other public sector organisations bring deep knowledge of their services, users and obligations. That knowledge can be strengthened by commercial delivery experience gained from solving comparable problems across different organisations: what delivered value, what implementation really took and which promises proved unrealistic.
When requirements are shaped mainly by existing processes and previous programmes, their limitations can carry into the next investment. AI can end up accelerating the way work already happens, while the opportunity to improve the service goes unexplored.
At BeInspired, we believe public bodies, primes and specialist partners need to bring their experience together before the requirement is fixed.
Our starting point
Public bodies gain a broader basis for judging what is possible and worth commissioning. Delivery partners gain the operational understanding they need to make credible commitments.
The scale of investment makes that exchange increasingly important. By early August, UK public bodies had awarded £1.41 billion in AI-related contracts during 2026, already exceeding the total recorded across 2025, according to Tussell. These are contract awards, including wider programmes with an AI component, rather than expenditure solely on AI technology. [1]
The challenge is turning that commitment into better services.
Knowing what good looks like
Broader delivery experience gives commissioning teams a basis for comparison. It helps them assess the ambition of a proposal as well as its feasibility.
What has worked in a comparable setting? What preparation did it require? Which costs emerged during implementation? How was the improvement measured?
The OECD’s Digital Government Outlook 2026 highlights the expertise governments need to define AI requirements, assess suppliers and retain effective oversight. [2] External experience can strengthen that capability, provided the public body retains the understanding and authority to interrogate advice and make informed decisions.
Suppliers have a responsibility here too. A successful approach elsewhere is a useful reference, but its relevance depends on the service, its users and its constraints. The most useful evidence explains the conditions behind the result: the process changes, integration, operational effort and costs that made it possible.
That gives both sides something concrete to work with. Commissioning teams can compare approaches on a more informed basis. Primes and specialists can expose assumptions before they become delivery commitments.
Understand the work before defining the solution
A faster task does not automatically produce a better service.
An application might take seconds to summarise and still wait weeks for missing evidence, another team’s assessment or an authorised decision. The saving is real. Its value depends on what happens next.
This is why process intelligence and process design belong in the commissioning conversation.
Where suitable operational data exists, process mining can reveal where work waits, repeats or generates rework. Staff and user research help explain why, including the workarounds that may never appear in the documented process.
Together, they provide evidence for deciding where AI, intelligent automation, integration or a redesigned process could make the greatest difference.
That can materially change the brief. The priority may be improving information at the point of entry, connecting two systems or changing an approval route. Accelerating one stage might simply move the queue downstream.
This is where operational knowledge and wider delivery experience work together. One explains why the service operates as it does. The other helps challenge assumptions about how it could work.
For public bodies and primes, the useful test is the same: if the proposed intervention works exactly as promised, what improves for the service and its users?
Make the full delivery requirement visible
Capgemini’s Data and AI in government 2026 report finds that only 28% of public sector organisations surveyed globally report successfully scaling their AI initiatives. Data quality, legacy systems, governance and workforce readiness remain significant challenges. [3]
Those challenges translate into substantive delivery work.
An AI service using information from several systems needs data engineering, integration and clear ownership. A tool supporting decisions needs evaluation, assurance and a route for handling uncertainty. Wider adoption requires changes to processes, responsibilities and support.
These elements need to be scoped, costed and owned within the commission. Otherwise, they emerge later as dependencies nobody has properly planned or funded.
The same discipline applies to benefits. Time saved on an individual task should connect to the outcome the organisation is paying to improve: resolution time, accuracy, repeat contact, cost per completed case or the experience of the service user.
If time savings are expected to reduce a backlog, what changes will allow that capacity to be used where it is needed? Who is responsible for making those changes?
The BeInspired perspective
At BeInspired, our view is that the whole service should be the basis for judging value.
That gives the client and its delivery partners a shared outcome against which to assess progress.
That shared understanding needs to survive beyond commissioning. During delivery, teams need access to operational staff, clear ownership of decisions and a way to resolve dependencies together. Knowledge transfer should be part of the work, so the public body can operate and improve the service with confidence after implementation.
Make SME participation part of that design
Public sector SME commitments create an opportunity to widen the expertise involved.
Alongside departmental targets for direct SME spending, buyers are encouraging participation through major suppliers. DSIT’s March 2026 SME Action Plan, for example, includes targeted SME engagement for AI procurements and requests for prime suppliers to consider SME opportunities within specified major programme supply chains. [4]
For public bodies and primes, the strongest starting point is the work.
Where would specialist knowledge improve the commission? Which workstream needs additional depth? What expertise would help challenge an assumption or resolve a dependency?
That might lead to an SME helping establish the process evidence behind a requirement, prepare data, engineer an integration or support service transition. Involving that expertise while the approach is being shaped gives it room to influence the outcome.
The contribution needs to be clear, with the access and accountability to deliver it. That makes SME participation a practical part of the programme’s capability and connects supply-chain commitments with the outcomes being commissioned.
Commission the change you need
AI investment creates an opportunity to rethink how services work. The quality of the requirement will determine how much of that opportunity a programme can pursue.
At BeInspired, we believe bringing operational knowledge and commercial delivery experience together early is essential to making that requirement stronger. It gives public bodies the perspective to commission with confidence and delivery partners a sound basis for the commitments they make.
We bring together specialist consultants and multidisciplinary teams across Technology, Transformation and Growth, working with public sector organisations, systems integrators and prime contractors.
If you are shaping a requirement, preparing a bid or addressing a delivery challenge, we would welcome a conversation.
Specialist expertise. New perspectives. Proven delivery capability.
SME. Without compromise.
Sources
- [1] Tussell: UK Public Sector AI Procurement Tracker, August 2026. Contract awards through 3 August 2026.
- [2] OECD: Digital Government Outlook 2026, Adopting and governing AI in government, June 2026.
- [3] Capgemini Research Institute: Data and AI in government 2026. Global public sector research.
- [4] Department for Science, Innovation and Technology: SME Action Plan 2025–2028, published March 2026.