900 staff and 42 branches is too much surface area for one workshop. This plan stages the rollout in four phases over roughly 12 months — starting with the 20 people who set policy, ending with branch-level pilots that pay for themselves.
Managers field the questions: "What does the library want us to do with AI?" If they get trained after staff, every conversation in between is a guess. Phase 2 gives them a script before the questions start.
CHPL's frontline staff don't sit at desks. The survey is one screen scroll on a phone, 6 minutes max. Training videos auto-play on mute with captions so they work in a quiet workroom. Live sessions are 25 minutes, scheduled during shift overlap windows.
A representative sample of ~300 gives the same statistical signal as 815 respondents — and avoids survey fatigue with staff who already get pulled in many directions. We can always re-survey a branch if its results look anomalous.
| Phase 1 · Execs | Phase 2 · Managers | Phase 3 · Staff | Phase 4 · Pilots | |
|---|---|---|---|---|
| People | 20 | 65 | ~815 (sampled) | — |
| Survey | Full, named | Full, by region | Anonymous, sampled | — |
| Format | 60-min workshop | 2-hr workshop ×3 regions | Async video + monthly lunch & learn | Workflow build sprints |
| Delivery | In person | In person + hybrid | Mobile-first | In branch |
| Duration | ~30 days | ~60 days | ~120 days | Ongoing |
| Output | Vocab + principles | Region readiness map | Branch index + champions | Working pilots, ROI |
Can we land the Executive Foundations workshop before your October/November staff development days? That gives Phase 2 enough runway to ride that momentum.
For Phase 1, who sends the link to executives — you two, or the CEO/Director? Attribution affects response rate.
Do all 815 staff have a CHPL-issued device, or do many use personal phones? That changes how we deliver Phase 3 content.
Who would own the scheduling pilot internally? We'd build alongside that person — they become the local expert, not us.