Skip to main content
The diagnostic should create a decision, not simply a report. If the evidence supports action, the natural next question is: Which opportunity should the client implement first, and who should deliver it?

​
Build an implementation backlog

For each opportunity record:
  • desired business outcome;
  • current-state evidence;
  • proposed intervention;
  • owner;
  • scope;
  • technical dependencies;
  • process dependencies;
  • data requirements;
  • human oversight;
  • implementation effort;
  • expected value;
  • success KPI.

​
Select the first project

Prefer a project with a strong combination of value, evidence, contained scope and a credible path to production. Avoid choosing a flashy initiative purely because it demonstrates sophisticated AI.

​
Decide delivery ownership

You may:
  • implement the work yourself;
  • use a specialist partner;
  • work alongside the client’s internal team;
  • recommend a vendor and remain as advisor;
  • stop after diagnosis if implementation falls outside your capability.
Credibility increases when you are explicit about your limits.

​
Transition talk track

Of the opportunities we identified, we recommend starting with X because it has the strongest combination of evidence, expected value and manageable implementation complexity. The next step would be to define the production scope, dependencies, success measures and delivery ownership.

​
Keep diagnosis and implementation commercially distinct

Unless explicitly included, detailed production design and build work should be a separate scope and fee.