Key takeaways
- A roadmap is not a list but a justified sequence.
- Difficulty scoring includes the process owner's willingness.
- The high benefit–low difficulty quadrant forms the first wave.
- Without quarterly review a roadmap ages within six months.
Start with a maturity assessment
Current state is mapped across five axes: data access, technical infrastructure, capability, governance and culture. Plans made without this typically start two steps ahead of where the organisation actually is and stall in the first quarter.
The output is not a score but a list of constraints: which use case is impossible for now, and because of what?
The use case portfolio
Short sessions with teams produce a decision inventory. The three properties sought are constant: decisions that repeat often, take time and are costly to get wrong.
The resulting list is typically 20–40 items. The list itself is not a plan; the plan is how the list is sequenced.
The value–effort matrix
Each use case is scored on two axes. On the value side sit hours saved, error reduction and revenue impact. On the effort side sit data access, integration, privacy sensitivity and the process owner's willingness.
That last item is the most often skipped and the most decisive. A technically easy use case with an unwilling owner moves slower than a technically hard one that is genuinely owned.
Quarterly milestones
The high benefit–low difficulty quadrant forms the first wave and usually becomes a 90-day starter programme. The high benefit–high difficulty quadrant spreads across the second and third quarters.
Each milestone carries a budget range, an accountable role and a measurable success criterion. The roadmap is reviewed quarterly and what is learned changes the sequence.
