Before tools, before automation, before AI. Name the operational pain the business owner actually feels on a Tuesday. The thing that keeps you in the day-to-day when you should be growing the company.
Most owners we audit spend 12 to 20 hours a week on work that has a system attached to it. They just haven't built the system yet.
— Steven Asoy, The Lean Operator
Clarity3/5
1 · No idea, I just react5 · Mapped, ready to delegate
Hours value3/5
1 · Not much5 · It would unlock real growth
Section 02
Your shop today
Rate the honest state of the operation, not the version you put in pitches. We're separating "I use ChatGPT" from "my business runs without me." Those are different problems.
ChatGPT in the browser is not a system. It's a tool. A system is when the work happens whether or not you open the laptop.
— operator's law
SOPs3/5
1 · It lives in my head5 · Every workflow has an SOP
Data state3/5
1 · Spreadsheets and DMs5 · One source of truth, queryable
Stack integration3/5
1 · Copy-paste between apps5 · Automated, two-way sync
Org clarity3/5
1 · I run everything5 · Roles, owners, and limits are clear
Permission to experiment3/5
1 · Scared to touch it5 · Sandbox + clear red lines
Section 03
Find the bottleneck
"Use AI in our business" is not a project. "Auto-draft Monday client reports from this week's data" is. Get specific about where hours bleed.
The owner who can name the workflow, the owner, and the hours saved already has a project. The owner who says "AI for ops" has a feeling.
— diagnostic principle
Definition3/5
1 · Vibes only5 · Scoped with hours saved + owner
Business value3/5
1 · I'm guessing5 · I can show the math
Section 04
Who's holding the rope
"Build systems, not teams" doesn't mean fire everyone. It means decide which work is system-shaped and which work still needs a human, then staff accordingly. Most owners over-hire because they never separated the two.
Every role you almost hired this year is a system you didn't build.
— operator's tradeoff
Data privacy, client contracts, access controls. Nobody wants to think about it. Then a tool you trialed leaks a client deck and the conversation gets very expensive, very fast.
The point isn't to build like an enterprise. The point is to know which two or three rules you can't break, so the rest of the system has room to move fast.
— governance for small shops
AI policy3/5
1 · No rules, just vibes5 · One-page policy, team has read it
Access control3/5
1 · Everyone has everything5 · Role-based, least privilege
Section 06
What you already pay for
Most owners overestimate what they need to buy. They already pay for half the stack. The win is connecting it. Audit what's running, who actually uses it, and what's a ghost subscription.
Before we add a tool, we look at what's already on the bill. The cheapest AI win is usually the integration you haven't built yet.
— audit principle
Integration3/5
1 · Manual everything5 · Automated, two-way
Custom builds3/5
1 · Only off-the-shelf5 · Workflows we built, we own
Section 07
The 21-day build sprint
A 90-day roadmap nobody runs is a deck. Twenty-one days is the longest window an owner stays bought in. Pick one workflow, name the owner, define what "shipped" looks like.
If you can't ship one workflow in 21 days, the problem isn't AI. The problem is scope.
— build sprint rule
Execution risk3/5
1 · Will get bumped by week 25 · Calendar blocked, owner committed
Section 08
Owner's call sheet
One page. The version you read to yourself on Monday morning to remember what you're trying to do. Not a board deck. Not a plan you'll forget. The call.
Owners don't need a strategy. They need a decision. This page is the decision.
— call sheet principle
What happens next
You finish the audit. We turn your answers into a tight findings report and a 21-day build sprint scoped to your business.