Start with the outcome, not the tool.
Most AI advice starts with what the technology can do, then looks for places to use it. I work the other way. First we agree what your business needs to achieve and find what's holding it back. Only then do we ask where AI genuinely helps.
-
01Outcome
Agree what success looks like.
Define the result you're after (more capacity, growth, or quality or less risk) in measurable terms. Set a timeframe and note the constraints that matter. AI doesn't come up yet.
-
02Understand
Find the real bottleneck.
Map how the work actually flows today: people, processes, systems, data, and any AI already in use. Record a baseline and identify the constraint that limits the outcome. Often it's data or process, not a missing tool.
-
03Reimagine
Explore what could be different.
Look at options on three levels: do the current task better, redesign the workflow, or rethink what the business offers. Every idea is a hypothesis to test, not a promise.
-
04Prioritise
Choose, with guardrails.
Weigh value, feasibility, time to value, cost, and how ready your team is. Privacy, security, and quality risk is a gate, not a score: if it can't be managed, the idea doesn't proceed. You leave knowing what to do now, what to explore next, and what to leave alone.
-
05Act
Prove it, then decide.
Run a small, controlled pilot with a named owner, human review, and clear conditions for stopping. Measure it against the baseline, then scale it, change it, or stop it. Then look again, because fixing one bottleneck reveals the next.
The method is the same for a ten-person firm after a quick win and for a board working through AI strategy and governance. The depth changes, not the approach.
Start a conversation