TL;DR: If any readiness lens is a hard 0, don’t start building—fix the gap first. This post gives you a fast 6‑lens check and a mini scoring example you can run today.
The 6 lenses (fast and honest)
1) Process clarity — AS‑IS documented, stable, and agreed?
2) Data fitness — Inputs complete, structured, and accessible (no screenshots-as-data)?
3) Tech foundations — Environments, access, APIs, logging, and monitoring exist?
4) People & skills — SMEs available, product owner engaged, dev/test capacity in place?
5) Governance — Change, risk, security and sign‑offs defined?
6) Value case — Benefits beyond FTE (quality, cycle time, capacity, compliance) quantified?
How to score in 5 minutes
Give each lens 0–3: 0=unknown, 1=weak, 2=workable, 3=strong. If any lens is 0, pause. If the average < 2, start with stabilize‑first actions rather than a build.
Mini example
| Lens | Score (0–3) | Notes |
|---|---|---|
| Process clarity | 2 | We have a rough BPMN, exceptions not documented. |
| Data fitness | 1 | Email attachments vary; duplicates. |
| Tech foundations | 2 | No API; UI access OK; logging TBD. |
| People & skills | 3 | PO + SME committed; dev time booked. |
| Governance | 1 | No risk log yet; security review pending. |
| Value case | 2 | Cycle time and accuracy improvements expected. |
Fix‑first backlog from the scores
Turn low scores into actions:
- [ ] Run a 2‑hour exception‑harvest workshop with SMEs.
- [ ] Create a sample input pack (10 real cases incl. edge cases).
- [ ] Set up a risk log with owners + weekly review rhythm.
- [ ] Draft a telemetry plan (what to log from day 1).
- [ ] Write a 1‑page value case with baselines and target deltas.
Red flags (pause here)
- [ ] Process owner can’t name success metrics.
- [ ] Data only exists as screenshots/hand‑entered text.
- [ ] Security won’t allow non‑prod access or test data.
- [ ] No single accountable sponsor for sign‑off.
- [ ] ‘Build it and we’ll find the value later.’
Make the call
If the average is ≥2.3 and no zeros, green‑light a tiny pilot (2–4 weeks) with a clear success metric. If not, invest 1–2 weeks in the fix‑first backlog—then reassess.
