How to know your business is ready for automation

A lot of automation projects don't fail because the technology was wrong. They fail because the business wasn't actually ready, and nobody said so before the contract was signed.

I've watched this pattern from both sides: as an operator who paid for it, and now as a consultant who refuses to. There are five readiness gates almost every successful AI or automation engagement passes through before the first line of code gets written. Miss one and you'll spend twice the budget for half the result.

This post walks through all five — with the scoring system we actually use with clients. By the end you'll know whether your business is ready, almost ready, or whether the right next move is foundation work first.

(There's also a free 4-page version of this you can download as a PDF — the AI Readiness Checklist — which is what we send every new newsletter subscriber. Same framework, printable, designed to do with your team.)

Why "are we ready" is the right question

Most businesses ask: "should we do AI?" That's the wrong question, because the answer is almost always yes — but the when is what determines whether you waste $20,000 or earn back $200,000.

The right question is: are we ready for AI to amplify what we already are?

AI is a multiplier. If your processes are clear and your team is aligned, AI multiplies that into faster, cheaper, more consistent work. If your processes are vague and your team is firefighting, AI multiplies the chaos. Same software, opposite results.

The five readiness gates are the things that determine which side of that line you're on.

The five gates

1. Strategy — do you know what problem you're actually solving?

This sounds insultingly basic. It is also the gate most companies fail.

The reliable pattern: a business says, "we want AI." When you press for what specifically — what process, whose week gets better, what number should move — the conversation thins out. "We just don't want to fall behind." "Our competitors are using it." "Our board asked."

That's not a problem statement. That's anxiety.

You're ready on the strategy gate when you can finish all of these sentences in plain English:

• "The specific problem we want AI to solve is _______________."
• "If we solved it, the number that would change is _______________, by about _____%."
• "The team most affected by this is _______________."
• "We've ruled out solving it without AI because _______________."

If you can't finish one of those, don't write the AI check yet. Spend a week with your team finishing them. That alone is more valuable than most "AI strategy" engagements.

2. Data — is your business legible?

You don't need a data lake. You need your business data to be:

• Digital. If a process is captured only on paper or in a senior person's head, AI can't help yet.
• Findable. Could your office admin pull a list of the last 50 jobs, with cost, duration, and customer, in 10 minutes? If no — gate failed.
• Correctable. When a number is wrong, is it obvious where to go to fix it? Or does fixing one thing break three others?

Most SMBs are at maybe 60% data readiness. That's fine — perfect data is not the bar. The bar is "the data exists somewhere I can point to, even if it needs cleanup." Cleanup is part of the engagement. Non-existent data is not.

3. Workflows — do the processes you'd automate actually exist?

You can't automate what isn't there.

I once watched a client try to automate "the way we approve POs," only to discover that there were four different unwritten approval flows depending on who was asking, what time of day it was, and which manager was traveling. Three months and a substantial bill later, the result wasn't an automated PO process. It was a standardized PO process with no automation yet.

Which was actually fine — the standardization was the win, and they got to automation a year later. But they paid AI prices for what was really an operations engagement.

You're ready on workflow when:

• The process you'd automate has at least a back-of-napkin diagram somewhere.
• You can describe the inputs, outputs, and the 4–7 main steps without making it up on the spot.
• The team doing the work today is open to the workflow changing.

4. Team — is there a champion, and 5 hours a week?

Every successful build has the same shape on the client side:

• One champion. Not "the team." One person whose name goes next to it in the directory, whose calendar shows weekly time blocked, whose performance review next year will reflect whether this thing landed.
• 5–10 hours a week of that champion's time, for the duration of the build (typically 4–8 weeks).
• Leadership cover. The champion has explicit air cover from the owner or executive team to push back on competing priorities for the duration.

Without those three things, the prettiest technical build still fails. With them, even a clunky first build usually succeeds, because the champion fixes things on the fly.

If you can't name your champion right now — or if the only candidate is already stretched across three other initiatives — your business isn't ready. Fix this first.

5. Risk — have you thought through the failure modes?

AI fails differently than software. It hallucinates, drifts, and degrades silently. A regular bug crashes loudly. An AI bug confidently gives you the wrong answer for three weeks before someone notices.

You're ready on the risk gate when:

• You've named the top 2–3 ways this can fail badly.
• For each, there's a human-in-the-loop check or a circuit breaker.
• You've reviewed the privacy / compliance / IP implications for your industry. (Healthcare, legal, finance: the bar is higher.)
• You've talked to clients or customers about AI use if it touches them.

This isn't paranoia. It's how every grown-up business handles a new capability.

The honest scoring

For each gate, score yourself 0–4:

• 0 — We haven't thought about it.
• 1 — We've talked about it; nothing concrete.
• 2 — Partially in place; significant gaps.
• 3 — Mostly in place; minor gaps.
• 4 — Solidly in place; we're confident here.

Add the five scores. Total out of 20.

Score	What to do next
16–20	You're ready. Pick the highest-leverage use case and scope a build. The Five Question Filter from the construction post is a good starting point.
10–15	The Bottleneck Audit is the right move. One week, $2,500. You'll know exactly which gate is your real bottleneck and have a prioritized fix list to clear it.
5–9	The Bottleneck Audit aimed at your weakest gates. One week, $2,500. You'll leave with a prioritized list of foundational ops work — workflow gaps, data hygiene, role clarity — to do before any build, plus the automation worth investing in once those are clean.
Under 5	Don't buy AI yet. The money you'd spend on a botched build buys six months of focused operations work that materially changes the company. Start there.

Most SMBs we talk to land in the 10–15 band. That's the sweet spot for a Sprint. You have real conviction, you know the business, and a focused two weeks of outside attention can identify the precise gap and the precise first build that'll move the needle.

What we actually do in a Sprint, when readiness is the bottleneck

I'll be transparent about what The Bottleneck Audit looks like when readiness is the issue, not just use-case prioritization:

• Week 1: stakeholder interviews, workflow mapping, gate-by-gate scoring with your team. We identify which of the five gates is actually blocking you and quantify the gap.
• Week 2: prioritized roadmap. Top 3 use cases ranked by impact × readiness. Effort and cost estimates for each. Recommendations for any foundation work that needs to come first.
• Deliverable: a short, defensible document you can show your board, plus a 60-minute readout. You leave the engagement knowing exactly what to do next, in what order, and what it should cost.

The point of the Sprint is to not sell you a big build you aren't ready for. About a third of our Sprints end with: "do these two operations things first, then come back in three months for the build." That's a feature, not a bug.

The bottom line

You don't need to be perfect at all five gates. You need to be honest about where you are. The companies that succeed with AI aren't the ones with the best technology — they're the ones who knew what they were ready for, picked the right first build, and let the wins compound.

If you want to go through this with your team, download the AI Readiness Checklist — same framework as above, in a printable 4-page PDF, free.

If you want a senior outside hand to run the diagnostic with you, the Bottleneck Audit is one week and $2,500. Either way, start with the gates. The technology is the easy part.