What 8 impact organisations taught us about AI adoption
Before the first mentorship session, we asked every organisation in our cohort where AI meets their daily work. Eight organisations is a small sample, so read everything below as a promising signal, not proof.
Eight themes kept coming up
How many of the 8 organisations raised each theme in their intake and early mentorship sessions. A theme only counts when an organisation named it explicitly.
Honest observations from the first weeks
AI gets you to 90%. The last 10% is still handwork
Drafts, summaries and translations arrive fast. Making them right for your audience still takes a person who knows the work.
The blocker is rarely the technology
Tools are the easy part. Time, ownership and permission to experiment are what actually decide whether anything sticks.
Trust starts with data, not features
Teams engaged only after they understood exactly what happens to their data. Skipping that conversation costs you the whole programme.
Small and boring beats ambitious and abandoned
One working tool for a weekly chore outperforms a grand roadmap. Every organisation that shipped something small kept using it.
Tallied from intake surveys, mentor briefings and session notes across the 8 organisations of cohort 1, plus an employee survey (n=31), conducted before and during the first mentorship weeks. n=8 at organisation level: a promising signal, not proof yet. All responses are anonymised; no individual organisation’s answers are identifiable in the data above.
