← Back to Courses
PRO 30 min read

Lesson 3 — The 30-Minute Weekly Ops Review

Turn your week's data — sales, traffic, customer messages — into 5 decisions in 30 minutes. A repeatable AI workflow.

Course Small Business · Lesson 3 of 4

Most small-business owners don’t do a weekly review because it’s a chore. AI makes it a 30-minute workflow that produces real decisions, not just numbers.

By the end of this lesson, you’ll have a weekly review template you can run every Monday morning to know:

The data you need (every week)

You don’t need a fancy dashboard. You need 4 things in any format AI can read:

  1. Revenue/sales — last week’s number, plus the previous week’s
  2. Top traffic sources — where customers came from (Google, Instagram, referral, etc.)
  3. Customer messages — count + the gist of the most common ones
  4. Anything you tried — a campaign, a promo, a new product, a content theme

Even rough numbers work. AI is good at finding patterns from imperfect data.

Step 1 — Drop the data into AI

Use Claude or ChatGPT Plus (both can read uploaded files; Claude has the bigger context window for spreadsheets).

You can paste numbers directly:

Here’s last week’s data:

REVENUE: - Last week: $X - Previous week: $Y - 4-week trend: $A → $B → $C → $X

TRAFFIC SOURCES (last week): - Google: 480 visitors - Instagram: 220 visitors - Direct: 95 visitors - […]

CUSTOMER MESSAGES (last week, ~12 total): - 4× shipping questions - 3× sizing questions - 2× refund requests - 1× compliment - 2× partnership inquiries

WHAT I TRIED: - New Instagram series (“Behind-the-build”) - Promo: 15% off for newsletter subscribers

Or upload a CSV/spreadsheet and ask AI to read it.

Step 2 — Run the analysis prompt

Here’s the prompt that turns the data into insight:

You are my operations analyst. I run a small business. Below is last week’s data.

Tell me:

1. What’s the headline? In one sentence — what kind of week was this? 2. What worked? Identify 2 specific things based on the data — be concrete, not generic. 3. What didn’t work? Identify 1–2 things that underperformed and why you think so. 4. The most interesting customer signal — what are customers telling me through their messages and behavior that I should act on? 5. One thing to test next week — based on what worked, what’s a small experiment worth running? 6. One thing to stop doing — based on what didn’t work, what should I quit?

Be direct. If you’re guessing, say so. Don’t tell me everything is great if it isn’t.

DATA: [paste from Step 1]

You’ll get a 1-page analysis. Read it. Disagree with it where you have better context — you know your business; AI is reading numbers.

Step 3 — Pull out 5 decisions

Here’s the magic step. After the analysis, prompt:

Based on the analysis above, give me 5 decisions to make this week. Format each as:

- DECISION: [what to do, in one sentence] - BY: [day of the week] - WHY: [1-line reason from the data] - HOW LONG: [time estimate]

Make them small enough to actually finish. Don’t suggest “redesign your website” — suggest “rewrite your homepage headline.” Tactical, not strategic.

You should get a list like:

  1. DECISION: Rewrite Instagram bio with new tagline matching this week’s top-performing post. BY: Monday. WHY: “Behind-the-build” series got 3× normal engagement. HOW LONG: 15 min.

  2. DECISION: Add a sizing FAQ to the product page. BY: Wednesday. WHY: 3× sizing questions in customer messages last week — same pattern 3 weeks running. HOW LONG: 30 min.

…etc.

These are your week’s plan. Calendar them.

Step 4 — The “what’s coming” forecast

Optional but powerful. Run this once at the end of the review:

Based on the past 4 weeks of data and patterns, what should I be paying attention to in the next 30 days? Highlight:

- 1 trend that’s emerging (good or bad) I might be missing - 1 risk I should watch - 1 opportunity I might not have seen

Be honest if the data isn’t enough to call something — say so.

This is the part that turns a review from “what happened” into “what’s about to happen.” Most small-business owners never do it because they’re too in the weeds. AI levels you up.

Step 5 — Save and stack

Each Monday, save the review. After 4 weeks, you have:

After 12 weeks, you’ll see seasonality, growth trends, and customer behavior that wasn’t visible week-by-week.

Run this prompt monthly:

Below are my last 4 weekly reviews. Tell me:

- The biggest pattern across the month - What I’m consistently doing well - What I’m consistently neglecting - One bigger move worth considering this month

[paste 4 weeks of reviews]

This is the closest thing a 1-person business has to a board of advisors.

Common failure modes

“My data is messy.” → That’s fine. Tell AI it’s messy. Paste what you have. It’ll work with imperfect data better than you’d think.

“The decisions are too vague.” → Add a constraint: “Decisions must each be doable in under 1 hour and have a measurable outcome.”

“I don’t believe AI’s interpretation.” → Trust your gut over AI when it conflicts. You know your business. AI is a fast, dumb-smart pattern-matcher; treat its output as a starting point, not the truth.

“I don’t have time for a weekly review.” → 30 minutes a week pays back hours. If 30 is too much, do a 15-minute version monthly. Some review beats no review.

The reusable template

Save this somewhere. Run it every Monday.

INPUTS:
- Revenue this week / last week
- Top 3 traffic sources
- Total customer messages + 1-line summary of patterns
- 1–2 things I tried

PROMPT 1: Run the analysis (Step 2)
PROMPT 2: Pull 5 decisions (Step 3)
PROMPT 3 (optional): Forecast next 30 days (Step 4)

OUTPUT: 5 calendar-ready actions for the week
TIME: 30 minutes

What you should have now

Where to next

You’ve finished AI for Small Business. Some good next moves:

← PREVIOUS
Customer Service That Sounds Like You

Get the next lesson

One email a week. The 5 things in AI worth knowing.