Lakeville AI MN
The 30-day plan

Your first 30 days using AI for business, on a sticky note.

No tools to buy. No subscriptions to start. You're not learning AI — you're learning where AI helps your business and where it backfires. Those are different skills, and the second one is where the money is.

The whole plan, on one sticky note:

  1. Days 1–2: Pick one annoying task.
  2. Days 3–5: Ask AI to do it — badly.
  3. Days 6–10: Add your context. Watch it get better.
  4. Days 11–15: Show it to one person you trust.
  5. Days 16–21: Pick a second task. Repeat.
  6. Days 22–30: Decide what stays. Drop the rest.
Days 1–2

i.Pick one annoying task. Just one.

Why this matters: most owners try to "use AI" in general — and get nowhere. You need a specific target. Not "marketing." Not "operations." A specific paper-cut that costs you 30 minutes every week.

The same kind of email you keep retyping. The weekly schedule. Drafting an estimate. Responding to the same five customer questions.

Write it on a sticky note. One sentence. That's the whole assignment for the first two days.

Common mistake Picking something too big. "Improve customer service" is not a task — it's a category. "Reply faster to 'do you do flat roofs?' emails" is a task.
Days 3–5

ii.Ask AI to do that task — badly.

Why this matters: bad output teaches you more than good output. You're calibrating, not deploying.

Open ChatGPT (or Claude, or Gemini — any of the free tiers is fine for now) and ask it to do the task. Don't try to make it perfect. The goal isn't a good answer — the goal is to see the gap between what AI gives you and what your business actually needs. That gap is where the learning is.

The exact promptI run a small [your kind of business] in Lakeville, Minnesota. I need to [the one task you picked]. Write me a first draft, and then tell me what you'd need to know about my business to make it better.

Read what comes back honestly. The places where it's wrong are the places where AI doesn't know your business yet. That's not a failure of the tool. That's just data. You'll fix it next week.

Days 6–10

iii.Add your context. Watch it get better.

Why this matters: this is the single biggest thing newcomers miss — and the difference between "AI doesn't work" and "AI works for me."

Take the same task. This time, paste in real context from your business. Your last three estimates. A few examples of your tone-of-voice from past emails. Your service area. The kind of customer you're talking to.

The free AI tools work way better when you give them your specific situation — and most owners never do.

The exact promptHere are three examples of how I usually write to a customer: [paste three short real examples] Now, in the same voice and using my actual approach, help me [the same task as before].

The output should feel noticeably more like you. If it doesn't, you didn't paste enough examples. Try again with more.

Days 11–15

iv.Show it to one person you trust.

Why this matters: you can't grade your own AI work. You're too close to it. One honest reader saves you from shipping something off-tone.

Not your whole team. Not LinkedIn. One person who would tell you the truth. A long-time employee, a spouse, a fellow business owner, a trusted customer.

Show them the output without telling them where it came from. Ask two questions:

The two questions "Would you have known a human didn't write this?"
"Where did it sound off?"

Their honest answer is worth more than every "10 AI prompts for business" article on the internet.

Days 16–21

v.Pick a second task. Repeat.

Why this matters: the value isn't in any single task — it's in the muscle of "is this a job for AI or not?" You can only build that muscle by doing it twice.

Now do the whole loop again with one more task. Not five more. One.

The point isn't to add more tools — the point is to build your own judgment of where AI helps and where it backfires. You're not learning AI. You're learning which parts of your business want AI in them and which parts don't. Those are very different skills, and the second one is where the money is.

Common mistake Picking a "harder" task to prove the AI is impressive. Pick another small, annoying, weekly thing. The skill compounds across small wins, not single big ones.
Days 22–30

vi.Decide what stays. Drop the rest.

Why this matters: a small thing you actually use beats a giant rollout you don't. Most "AI strategies" fail by being too ambitious on day one.

Most owners who do this honestly end up keeping one or two small things and dropping everything else. That is the right outcome. The "transformation" decks won't tell you that, but it's true.

Write down what you're keeping, what you're dropping, and why. That's your real AI strategy. It will fit on one sticky note. Most "AI strategies" with twenty pages are worth less than your sticky note.

Your AI strategy, day 30
  1. Keeping: ___________________
  2. Dropping: ___________________
  3. Why: ___________________
When you finish

Stuck somewhere in the 30 days?

That's normal. That's exactly what the working group is for. Show up to the next session, bring the specific thing you got stuck on, and we'll work it out together — with people who got stuck in the same place last month.

This is the part that doesn't fit on the sticky note. The community is the difference between owners who quit at day 12 and owners who finish.

See the support paths · Ready to go further? Read the build-it method.

When you're ready

Come to a session. Bring your sticky note.

We'll see what you learned and where you got stuck. The community helps each other through.

Hop on the list → Build it yourself → · When you need support →