Evidence, not opinion
The research behind the seats — where we take a position, we show the working. Nothing here is invented; every figure is attributed to a named, checkable source.
Problem before platform
The most expensive mistake in any transformation isn't the technology, the vendor, or the methodology — it's the sequencing. Choose a platform before the business problem is precisely defined, and the programme is at risk regardless of what the budget line says. This has been true of ERP rollouts and operating-model redesigns for decades. It's just easier to see right now in AI.
Gartner forecasts that 40% of agentic AI projects will be cancelled by the end of 2027. MIT's Project NANDA found that 95% of enterprise generative AI pilots show no measurable impact on the P&L. S&P Global's adoption data and Studio Graphene's 2026 UK research point the same direction. The common thread isn't the model or the vendor — it's organisations reaching for a platform before they've defined the problem precisely enough for anyone to know what success looks like. Every seat on this site exists to fix that sequencing problem, whether the platform in question is an AI model, an ERP system, or a new operating model.
Why 97% of organisations can't show AI value
Synthesised from Deloitte, PwC, McKinsey, Stanford HAI, Gartner, Forrester and more than a dozen further sources: 88% of organisations now use AI in at least one function. Only 3% describe themselves as fully satisfied with the return. 97% struggle to demonstrate clear business value from what they've built. 74% of all AI value created is captured by the top 20% of firms — the gap between AI leaders and everyone else is widening, not narrowing.
The barriers are consistent across the research: a formal AI strategy takes success rates from 37% to 80%, yet most mid-market organisations don't have one. Only 22% of UK businesses have provided AI-specific governance training. 42% of Fortune 500 AI projects were abandoned in 2025, and just 7% of organisations have fully scaled an initiative past pilot. None of this is a technology problem. It's a leadership and governance problem, at exactly the scale this practice exists to solve.
Agile Sourcing: the real numbers
Agile sourcing applies sprint-based, cross-functional working to procurement cycles that would traditionally take months. The published case evidence is genuinely strong: organisations following the Lean-Agile Procurement model have taken as little as four weeks to a signed ERP contract, and forty-five days to a competitively bid public-sector award. One documented case took a 9% savings target and delivered 15% instead.
The research base spans the Agile Business Consortium's Swiss Casinos Group and Air France KLM Martinair Cargo case studies, the LAP Alliance's published Zalando case, a CMS Medicare payment-systems modernisation, AGCO's pandemic-era supply chain response, and CIPS's own guidance on agile and crisis-ready procurement. The discipline that matters is knowing where it applies: agile sourcing is not the right tool for every category, and regulated or highly complex sourcing usually needs a more conventional process. Choosing correctly is as much a part of the method as running the sprint itself.
Explore by seat
Fractional Procurement
Where the agile sourcing research turns into a delivered engagement.
View the seatWant the research applied to your business?
Fifteen minutes tells us whether there's a real conversation to have.