CEO playbook

The 90-day AI workforce transition plan.

A pragmatic playbook for the first 90 days after running the AI Job Impact Calculator on your roster. Communication, augmentation, retraining — without losing trust.

Phase 01 · Discover

Days 0–14 — Discover the truth on the ground.

Before you communicate anything externally, score every role honestly and brief leadership and HR. The risk classification is workforce planning input, not a personnel decision.

  • Run Path B (company roster) on the full org. Decide who can see the report (suggest: CEO, CHRO, COO).
  • Brief managers in private on REPLACE / AUGMENT / UNCHANGED — no public commentary yet.
  • Identify the two highest-leverage AUGMENT roles to pilot AI tools on first.
  • Pick one operational metric per pilot (cycle time, error rate, customer effort score).

Phase 02 · Plan

Days 15–30 — Plan the public narrative.

Sequence the rollout so trust compounds. Inside-out beats top-down: managers brief their teams before the all-hands.

  • Manager toolkit: a 1-page brief per role with the score, classification, and the lateral retraining path.
  • Town-hall messaging: workforce planning + augmentation, never termination. The methodology is open.
  • Approve budget for AI literacy certification — universal, not opt-in.
  • Decide who owns AI ops longer term (model updates, vendor risk, evaluation).

Phase 03 · Reskill

Days 31–60 — Reskill, don’t replace.

The single biggest mistake here is treating REPLACE classification as a termination signal. The same engine surfaces three retraining paths per role.

  • For each REPLACE role, surface the lateral path and budget the upskilling program.
  • Build internal AI literacy certifications. Make them universal, not opt-in.
  • Reset performance metrics so AI-augmented output is measurable, not just hours worked.
  • Pair every retraining cohort with a sponsoring manager and a 60-day check-in.

Phase 04 · Measure

Days 61–90 — Measure, then commit.

By day 60 your pilots have data. Use it to make build vs buy decisions on the augmentation tooling, and publish the plan internally.

  • Review pilot KPIs against the operational metric you picked on day 14.
  • Decide build vs buy on the augmentation tooling that worked.
  • Build durable AI ops: who owns model updates, who handles vendor risk, who runs evals.
  • Publish the plan internally. Trust compounds when employees see the plan in writing.

Five traps to avoid

The pitfalls every CEO hits in the first 90 days.

  1. 01

    Treating REPLACE classification as a termination signal — it is workforce planning input, not a personnel decision.

  2. 02

    Buying tooling before scoping the job-task it accelerates — the pilot picks the tool, not the other way around.

  3. 03

    Skipping the augmentation pilot — the data you collect there decides the build-vs-buy call later.

  4. 04

    Letting AI literacy stay opt-in — the productivity gap between augmented and non-augmented workers is 2–5×.

  5. 05

    Forgetting that regulated edge-case humans are the brand promise — AI breaks the brand precisely where it gets cocky.

Start with the data

Score your roster.

The 90-day plan starts with an honest classification. Five minutes of paste and you have a board-ready PDF.

Run the calculator