Ask a boardroom about digital transformation and you'll hear ambition: new platforms, AI everywhere, a business reinvented for the next decade. Ask the same room eighteen months later and you'll often hear something quieter — a programme that consumed budget and produced slides, dashboards of "workstreams," and very little that customers or the P&L ever felt. The vision was rarely the problem. The roadmap was.
The statistic has barely moved in a decade: the majority of transformation programmes fail to deliver the value they set out to. And when you examine the ones that stall, the shape is depressingly consistent — a grand three-year plan, a big-bang cutover, a large vendor, and nothing in production for a very long time. By the time reality arrives, the money is spent and the appetite is gone.
A roadmap that actually ships is a different animal. Here's what separates it from the binder.
1. Sequence by value and risk, not by system
The failing roadmap is organised around systems: "Phase 1, replace the ERP; Phase 2, the CRM." The shipping roadmap is organised around outcomes, and starts with one high-value, self-contained slice you can deliver without betting the company. You're looking for the process that's painful, measurable, and reasonably bounded — the reconciliation backlog, the manual month-end, the quote that takes three days. Win there first.
2. Measure outcomes, not outputs
This is the discipline that keeps a programme honest. "Migrated to the cloud," "rolled out the new portal," "adopted AI" are outputs — activity, not value. "Cut the month-end close from nine days to three," "halved quote turnaround," "removed £X of recurring cost" are outcomes. Every step on the roadmap should name the business number it moves and capture the baseline before you build — the same rigour behind proving AI's ROI. No metric, no mandate.
3. Ship in thin vertical slices
The multi-year phase is where transformations go to die. The alternative is thin vertical slices — small, end-to-end pieces of real capability that reach production every few weeks rather than a "big reveal" at the end. Each slice de-risks the next and starts compounding value immediately. This is the same operating-model lesson behind closing the AI production gap: shipping something real, supervised, and measured beats a perfect plan that never launches.
4. Fund the next win from the last
Grand roadmaps demand a lump-sum bet before anything is proven. Shipping roadmaps run on momentum: one credited result funds — and builds the internal confidence behind — the next. It's the same logic as treating your estate like a balance sheet — spend where the return is proven, and let each win earn the mandate for the next. It also means you can stop cheaply if a bet isn't working, instead of discovering it two years and seven figures in.
5. Change how people work, not just the technology
The most common cause of a "successful" transformation delivering nothing is that nobody adopted it. New systems that people route around are pure cost. Adoption — training, incentives, redesigned ways of working — isn't a phase you bolt on at the end; it's part of every slice. Technology delivered is not the same as value realised, and the gap between them is change management.
The uncomfortable truth about "comprehensive"
A boil-the-ocean roadmap feels safer to a board precisely because it looks comprehensive — every system, every workstream, a plan for everything. It is actually the riskiest option available, because it defers the moment of truth to the very end: you learn whether it works only after the budget is gone. Incremental delivery inverts that — you learn early, cheaply, and often, and you course-correct while it still costs little to do so. This is exactly the pragmatic, outcome-led approach behind our digital transformation work: a focused audit, a scalable blueprint, and 90-day sprints that put something real into production — not a slide deck.
The bottom line
Digital transformation doesn't fail for lack of ambition; it fails for lack of shipping. Pick one painful, valuable process. Put a measurable improvement into production in 90 days. Prove the number, then do it again — each win funding and de-risking the next. That's not a smaller ambition; it's the only version of the ambition that survives contact with reality.
Sitting on a transformation plan that hasn't shipped anything yet? Let's make it move.