AI Strategy
Insightful AI builds AI strategies for UK SMEs, charities, and public sector organisations. We start with your goals and constraints, not a technology wishlist. Where AI is not the right answer, we say so. Where it is, we sequence the work so you see results within 90 days and build from there.
How use cases are shortlisted
Value
Feasibility →
Market context
Most organisations that adopt AI without a plan waste money. Over 80% of AI projects fail, twice the rate of traditional IT projects (RAND Corporation). And 42% of UK companies abandoned most of their AI initiatives in 2025, up from 17% the year before (S&P Global, 2025).
Ready to talk?
30-minute call. No obligation.
Those numbers do not come from poor technology. They come from poor planning: tools bought without a clear problem to solve, pilots that never connect to the business case, and AI experiments that nobody governs, measures, or scales.
A strategy changes that. Not a slide deck about what AI could do in theory, but a working plan with timelines, costs, owners, and measurable outcomes tied to how your organisation actually operates.
Our approach follows the adoption lifecycle recommended by DSIT (the Department for Science, Innovation and Technology), the ICO, and the techUK/BridgeAI AI Adoption Framework. The same methodology the UK government uses, scaled for organisations your size.
A prioritised AI adoption roadmap with phased timelines, costs, projected returns, and governance requirements for each initiative. The roadmap is board-ready and includes quick wins you can act on within 90 days. You also receive an AI policy pack, a skills and training plan, and pilot designs for your highest-priority use cases.
The full deliverable set includes nine components, each aligned with UK government and industry body expectations:
This is not a report that sits on a shelf. Every phase produces working documents with named owners, and the first phase identifies wins you can act on within weeks.
Every AI strategy engagement at Insightful AI follows four phases, grounded in the DSIT 'Scan, Pilot, Scale' approach and the ICO AI and Data Protection Risk Toolkit. The full six-stage methodology is described on our how we work page.
We run stakeholder interviews and workshops to understand your goals, pain points, and constraints. We review your existing digital estate and data assets, including any informal AI use already happening. Candidate use cases are scored by value, risk, and feasibility. An initial ethics and risk scan rules out inappropriate uses before time is invested in them.
You receive a current state report, a prioritised shortlist of candidate use cases, and clear identification of any problems better solved without AI.
We assess AI readiness across leadership, culture, skills, processes, and infrastructure. Data readiness is assessed against GDS's four AI-ready data pillars: technical optimisation, data quality, organisational context, and legal compliance. Use cases are classified into risk tiers (green, amber, red) so governance requirements match the actual risk of each initiative.
You receive a readiness and risk assessment with maturity scoring, a data readiness assessment, and a refined use-case shortlist with go or no-go recommendations for each.
We co-create your AI vision and principles, run a prioritisation workshop, and build a 12 to 24 month phased roadmap. Quick wins are sequenced first (90-day pilots), followed by medium-term initiatives and longer-term capability building. The business case includes projected costs, returns, risks, and sensitivities.
You receive the strategy document, the phased roadmap with milestones and dependencies, and a board-ready business case.
We design your AI governance structure, including oversight responsibilities, escalation paths, and integration with your existing risk and compliance processes. We draft your AI policy pack (principles, acceptable use, procurement standards) and design one to two priority pilots with scope, success metrics, data flows, and evaluation criteria.
You receive the governance structure, the policy pack, a skills and training plan, and pilot design documents with monitoring and evaluation plans.
Each phase has a governance checkpoint where your senior responsible owner signs off before the next phase begins. Nothing proceeds without your approval.
From £1,500 ex VAT for the scoping workshop that starts every engagement.
All strategy work begins with a scoping workshop at £1,500 ex VAT. The workshop is a full day (on-site or remote) and produces a scoped proposal for the strategy engagement. Strategy engagements are then priced as fixed-fee projects based on the workshop findings, so you know the total cost before committing.
This is a fraction of the cost of failed tool purchases and abandoned projects. RAND Corporation research puts AI project failure rates above 80%. A single failed implementation typically costs more than the strategy that would have prevented it.
Market benchmarks for comparable AI strategy engagements in the UK range from £3,000 to £50,000 depending on scope. Insightful AI's pricing is designed for SMEs and charities, not enterprise budgets.
For charities managing squeezed finances, a proportionate strategy identifies the use cases that protect your mission, not the ones that sound impressive in a funding bid. For councils and public bodies, the roadmap produces the documentation procurement and oversight teams require.
Use-case prioritisation is where the strategy either earns its keep or becomes another unused document. We score every candidate use case against three criteria: value to your organisation, feasibility given your current data and skills, and risk (regulatory, ethical, and operational). Use cases that score poorly on any criterion are deferred or excluded, with the reasoning documented.
71% of UK businesses have not identified a use case relevant to their organisation (DSIT, 2026). That is often a discovery problem, not an evidence problem. The stakeholder workshops and process mapping in Phase 1 surface opportunities that internal teams miss because they are too close to their own routines.
Where a use case is better solved without AI, we say so. That distinction saves money and builds trust with your board.
No. Data readiness is assessed as part of the strategy, not as a prerequisite to it. Where data quality or availability is a barrier, the roadmap includes data preparation as a sequenced step. Waiting for a perfect data estate before thinking about AI means waiting indefinitely.
The Phase 2 data readiness assessment is structured around GDS's four AI-ready data pillars. It identifies specific gaps and builds remediation into the roadmap with realistic timescales. Where specialist data work is needed, our data preparation and advisory service picks up from the strategy findings.
Governance is built into every phase of the strategy, not added at the end. The ICO (Information Commissioner's Office) expects organisations to demonstrate accountability for AI use. DSIT guidance, the Data (Use and Access) Act 2025, and the EU AI Act's requirements (transparency since August 2026, high-risk from December 2027) all apply regardless of organisation size.
In Phase 1, we run an initial ethics and risk scan against DSIT's cross-sector AI principles. Phase 2 classifies use cases into risk tiers and initiates DPIAs (Data Protection Impact Assessments) for higher-risk cases. Phase 4 produces your governance structure, AI policy pack, and acceptable use documentation. The result is a strategy that meets current regulatory expectations and positions you for what comes next.
For a deeper look at how we approach governance across all engagements, see our AI ethics and governance service page and our responsible AI position.
Insightful AI works with SMEs, charities, and public sector organisations across the UK, with a particular concentration in the North West of England. Our co-founders bring distinct strengths: Ben Sefton leads on strategy, governance, and client relationships, and serves as Chief AI Officer at Cheshire Community Foundation. Kane Lukassen leads on technical delivery, covering data science, software development, and AI system architecture.
We have delivered work for the National Emergencies Trust, Cheshire Community Foundation, Cumbria Community Foundation, and Book Aid International. Ben has presented on AI to the National Crime Agency, the National Police Chiefs' Council, Lancashire Police, and City of London Police.
The methodology we follow is not something we invented. It is grounded in the same frameworks used across UK government and industry: DSIT, GDS (Government Digital Service), the ICO, and the techUK/BridgeAI AI Adoption Framework. That means the outputs are structured in the formats your board, trustees, or procurement team already expect.
The strategy is designed to produce results in stages. The first 90 days focus on quick wins, use cases that can be piloted or implemented with your existing tools and data. Medium-term initiatives run over 6 to 12 months. Longer-term capability building and scaling extends to 24 months.
DSIT research found that 71% of UK businesses considered AI for about a year before deploying (DSIT, 2026). External perspective accelerates that timeline. The scoping workshop alone surfaces opportunities that internal teams have been circling for months.
76% of charities are using AI tools but only 2% strategically (Charity Digital Skills Report, 2025). A proportionate strategy for a 30-person charity looks very different from one built for a FTSE 100, and it should. The timescales and investment levels are matched to what your organisation can absorb.
| Your situation | What you need | What we deliver |
|---|---|---|
| Using AI ad-hoc with no coordination | A plan that connects AI to your actual goals | Prioritised roadmap with timelines, costs, and quick wins identified for the first 90 days |
| Bought tools but not seeing results | An honest assessment of what went wrong and what to do next | Independent review, reprioritisation, and a realistic path forward |
| Board or trustees asking about AI and you cannot answer confidently | A position you can defend with evidence | Board-ready documentation with governance, risk, and ROI projections |
| Charity or public body with limited budget | A proportionate strategy that fits your resources | Phased plan starting with the highest-impact, lowest-cost opportunities |
| Staff experimenting with AI tools and no one governing it | Visibility and control before something goes wrong | Shadow AI assessment, acceptable use policy, and governed use-case pipeline |
Myth
'AI strategy means buying the right tools.'
What actually happens
Many organisations approach AI by purchasing tools (ChatGPT, Copilot, CRM add-ons) and expecting change. Without a clear problem definition or roadmap, these purchases produce little. We have seen organisations spend six-figure sums on AI projects that deliver nothing because they started from technology, not business problems.
Myth
'A strategy is a one-off workshop.'
What actually happens
Workshops and roadmaps are the start, not the end. The strategy connects to funded pilots, implementation through services like AI at Work and AI automation, and ongoing governance. Insightful AI is not a firm that writes a strategy and hands it off. We also build and implement.
Myth
'We're too small for an AI strategy.'
What actually happens
FSB research found 26% of small businesses do not believe AI is appropriate for their business. Over 90% of the increase in identified UK AI companies between 2023 and 2024 came from SMEs (DSIT, 2024). A proportionate strategy for a 50-person professional services firm is a different document from one written for a FTSE 100. It should be.
Myth
'Governance can wait until we've proved AI works.'
What actually happens
UK AI governance is not optional. The ICO expects organisations to demonstrate accountability for AI use regardless of size. DSIT guidance, the Data (Use and Access) Act 2025, and the EU AI Act all apply. Governance is built into the strategy from Phase 1, not bolted on after.
Myth
'We just need training, not strategy.'
What actually happens
Training builds awareness. Strategy gives that awareness somewhere to go. Without a plan, trained staff experiment in isolation, creating the shadow AI problem: 70% of employees using AI have not told their employer. Strategy connects training to governed use cases with measurable impact. Our AI training programmes are designed to run alongside or after strategy work.
The strategy connects directly to implementation. Where the roadmap identifies task-level wins, our AI at Work service builds working solutions. Where it surfaces automation opportunities, AI automation and agentic AI workflow design take over. Where custom software is needed, AI software development delivers it. And 6 to 12 months after implementation, AI impact and ROI auditing measures whether the roadmap delivered what it projected.
One team, from strategy through to working results. No hand-off to a different firm.
Healey's growth speech restated most of its AI content. The £100 million competition is for AI companies. For everyone else, the official statistics point at a gap between what staff already use and what employers know about.
Read moreAI has become an essential driver of business success in 2025, with more companies integrating artificial intelligence into core operations. Studies indicate that over 70% of global businesses leverage AI for automation, predictive analytics, and decision-making.
Read moreAI leaders are calling for stronger controls on frontier systems, not a halt to everyday use. UK organisations should govern autonomy before granting it.
Read moreWhether you're just getting started or looking to scale, we'll help you find the right path, responsibly.