How we scope a 5-week readiness assessment

Five weeks sounds fast for something that ends in a roadmap leadership will fund. Here’s how we actually spend them.

Week 1: data and infrastructure audit. We look at what data actually exists, not what the org chart says should exist, and get honest about how far from clean it is.

Week 2: stakeholder interviews. We talk to the people doing the work the AI system would touch, not just the executives sponsoring the assessment. The gap between what leadership thinks a workflow looks like and what the team actually does with it is usually where the best use cases hide.

Week 3: use-case scoring. Every candidate use case gets scored against the same matrix — effort, risk, payback — so three departments’ competing proposals get compared on the same terms instead of whoever pitches loudest winning the budget.

Week 4: build-vs-buy and architecture sketching. For the top-ranked use cases, we sketch what building it would actually involve, and flag anywhere an off-the-shelf tool would get there faster and more cheaply than a custom build.

Week 5: the roadmap. A 12-month plan with budget bands per use case, sequenced by payback and dependency, presented to whoever needs to fund it.

The output isn’t a slide deck people file away. On the Northfield & Cole engagement, this exact process produced a roadmap that got the company from a stalled AI budget to three production use cases within six months — because the roadmap named a first move, not just a vision.