Strategy that ends in a scope you can build from

Product discovery

Two to three weeks with the engineers who would build the thing. The output is a scope, an architecture and a cost, which is a different artefact from a strategy deck and a more useful one to a board.

• The problem stated in terms somebody outside the room would recognise
• Constraints found early, since the ones discovered in month four are the expensive ones
• An architecture with the trade-offs written down and the rejected options named
• A cost and a sequence you can commit to, or a recommendation not to proceed

AI readiness

On AI work discovery has an extra output, and it comes first. Without a written definition of a good answer there is no way to tell an improvement from a regression, and the programme runs on opinion.

• A scoring rubric agreed with the people who will judge the output
• The data that would ground it located, with the permission model examined
• A pilot scoped so its result actually decides something
• The cost of running it at volume estimated before anyone commits to it

Estate assessment

For the organisations where the question is what to keep rather than what to build. An assessment is only worth commissioning if it ends in a sequence somebody can fund.

• What runs, what it costs and who depends on it, written down once
• Decommissioning treated as a dated phase rather than an aspiration
• Risk stated in terms of what breaks and who notices
• A wave plan where the first wave is small enough to prove the approach

How we staff it

The same senior people throughout, which is the arrangement the company was started on. There is no account layer between the first conversation and the people who would do the work.

• A senior engineer in the first call, joined by product when the question warrants it
• The people who scope the work are the people who build it
• A team that stays through delivery rather than handing over at kickoff
• An honest recommendation when the work is smaller than the engagement