Readiness · Checklist
AI Readiness Checklist for $2M–$50M Services Firms
Eight questions to answer before you green-light an AI build. Stages mapped; nothing blank.
Introduction
Most AI engagements stall before they ship — not because the tech is hard, but because no one wrote down what the firm actually does, in what order, with which tools. A readiness checklist fixes that. The eight items below map to the four stages of the Plumbline AI Readiness Audit, so the answers you give on day one are the same answers that become the ranked roadmap on day fourteen.
The checklist is generic on purpose. It is the same checklist for an eight-person consultancy, a forty-person wealth advisory, and a specialty construction firm with $20M in revenue. The dollar answers differ; the discipline does not.
Why the four-stage frame makes the checklist generic-portable
Each stage closes with a written deliverable: a readiness read, a ranked list of opportunities, a costed 90-day roadmap, and a monthly operating cadence. The checklist below is in that order. If you skip an item, the deliverable that follows it gets weaker, and the partner who reviews the audit will be the one to flag it. Working through the checklist today is a cheaper way to find the gap than a two-week engagement is.
The checklist
Eight items. One item per line.
Each item is short and imperative. Answer yes or no on the line, and write the dollar or hour number where the word "ranked" appears. The answers become the ranked list the audit ships in week two.
- Readiness
List every recurring internal workflow that runs on email, spreadsheets, or chat threads. One column, no commentary.
- Readiness
Inventory the SaaS tools each team touches in a paid week, and rank the ones with the most state (most fields, most edits, most accounts).
- Sequence
Pick the three workflows where a missed hour costs real money or a real customer. Settle on three — picking seven is the same as picking zero.
- Sequence
Score those three on a single axis — dollars or hours, not both. Ranked, not brainstormed; defensible to a board in two sentences each.
- Deploy
Decide the stack call per agent: custom code by default; an off-the-shelf automation tool only when the workload is genuinely a better fit for one.
- Deploy
Ship the first agent in your accounts, never in a contractor’s. Code you own is portable by construction; a hosted workspace is not.
- Operate
Lock a cadence: weekly agent health digest, monthly prompt retune, quarterly workflow scan. The numbers are arbitrary; the rhythm is not.
- Operate
Name the single owner who reads the digest — a real person, not a shared inbox. If no name fits, the cadence is not real yet.
What you walk away with after running this
If you can answer every item in writing, you already have the inputs the audit would otherwise spend two weeks pulling out of stakeholders. The readiness read becomes a one-day read instead; the ranked opportunities become a one-day rank instead. The dollar answer is the same — the time-to-decision shrinks. If you cannot answer every item honestly, the audit is exactly the diagnostic that fills the gap. Either way is a good outcome.
Keep reading
The other post in the pair.
The five (plus one) agent deployments the Plumbline AI Readiness Audit ranks highest across owner-led services firms — sized by hours saved, dollars added, or exposure reduced, and shipped in the firm’s own accounts.
← Read 5 AI Agents Every $2M–$50M Services Firm Should ConsiderTranslate the checklist into a ranked roadmap
Book a Readiness Audit. Walk away with a ranked list.
Two weeks of interviews, a toolchain scan, a ranked list of agent deployments sized by dollars or hours, and a costed 90-day roadmap. Fixed fee, dated deliverables, and a partner-led build path that runs from the audit forward.