Founder's Weekly Ops Review: 30 Minutes, 5 Questions, No Spreadsheets

If you're an owner whose weekly review has expanded to two hours of dashboard scrolling, three open spreadsheets, and zero decisions written down — you're not doing a review. You're doing data tourism. The fix is shorter, not longer.

Why does the typical weekly review fail?

Because it's been confused with a status meeting for an audience of one. The founder opens five tabs, reads numbers, feels informed, and writes nothing down. By Wednesday the "review" has produced no traceable decision, and by next Sunday it gets repeated.

Definition: Data tourism — the act of scrolling dashboards in lieu of a decision, producing the feeling of management without producing a management artifact.

A review that produces no artifact is not a review. It's a habit that looks like one.

What makes a 30-minute review work?

Three constraints. First, the timer is real — set 30 minutes, stop at 30 minutes. Second, the output is a single document that gets read at the start of next week's session. Third, every question forces a written sentence; "thought about it" doesn't count.

Definition: Review artifact — a written one-page output from a weekly review that is re-opened at the next week's session, creating a compounding audit trail of decisions and patterns.

The artifact is what turns 52 weekly reviews into one coherent year of management decisions. Without it, you're starting over every Sunday.

The 5 questions

Question 1: What did the Plan say, and what actually happened?

Open last week's review artifact. Read what you said you'd do. Write what actually happened. Don't soften it. The gap between Plan and Fact is the input to every other question.

Definition: Plan-Fact delta — the written difference between what last week's review said would happen and what did happen, used as the primary input for diagnosis.

If you haven't written a Plan, write a backwards one — "the implicit plan was X" — and start the cadence properly next week.

Question 2: What surprised me this week?

Three to five surprises, written as bullet points. A surprise is anything you couldn't have predicted on Sunday morning. Bucket each one as either "external" (market, customer, supplier) or "internal" (team, process, infrastructure). Over four weeks, the internal/external ratio is the most useful number in this review.

A high internal-surprise rate means the company's internal state isn't legible to you. That's a leadership signal, not a logistics one.

Question 3: Where did I see people friction?

Not "who was a problem." Where in the org did two roles, two teams, or two priorities collide? Friction is a leading indicator of structural debt. It costs almost nothing this week and a quarter next quarter.

Write each friction point as a verb: "Sales and CS disagreed on onboarding scope for the Tier-2 segment." Then write the next-best action — usually a 30-minute conversation you've been postponing.

Question 4: What did cash try to tell me?

One paragraph. Cash signals are not just runway — they're collection speed, expense surprises, deal slippage, refund rate, churn-driven MRR changes. Most founders look at runway weekly and ignore the texture.

Definition: Cash texture — the qualitative pattern of cash movement (timing, source mix, slippage) underneath the headline runway number, often more diagnostic than the runway itself.

If something here surprises you, it goes back into Question 2 next week. Compounding pattern detection.

Question 5: What are next week's three priorities?

Three. Not five. Not "everything I'm doing." Three items that, if next Sunday's Fact column shows all three completed, you'd call the week a win. Write them as outputs, not activities — "Tier-2 pricing finalized with CFO" beats "work on pricing."

These three become the Plan that Question 1 will measure next week.

Copy/paste template

This goes in a single Notion page, Markdown doc, or A4 sheet. No spreadsheets.

WEEKLY REVIEW — [WEEK OF YYYY-MM-DD]
Time start: __:__ | Time end: __:__ (cap 30 min)

1. PLAN → FACT (read last week's "Next-week three")
- Plan 1: [...] | Fact: [...]
- Plan 2: [...] | Fact: [...]
- Plan 3: [...] | Fact: [...]

2. SURPRISES
- [...] | external / internal
- [...] | external / internal
- [...] | external / internal
Pattern check (4-week rolling): internal-to-external ratio: ___ / ___

3. PEOPLE FRICTION (verbs, not names)
- Friction: [...] | Next action: [...]
- Friction: [...] | Next action: [...]

4. CASH TEXTURE (one paragraph)
[...]

5. NEXT WEEK — TOP 3 OUTPUTS
- [Output, owner = me, by EOD Friday]
- [Output, owner = me, by EOD Friday]
- [Output, owner = me, by EOD Friday]

ONE-LINE WEEK NAME
This was the week of: [...]

The one-line week name at the bottom is the only memory aid that works at the 6-month horizon. "Week we finally killed the Tier-2 onboarding flow" is searchable. "Week of May 18" is not.

Tool tip (AIAdvisoryBoard.me): The weekly review only works if it sits on top of accurate daily Plan → Fact → Gap signal. Without that, Question 1 turns into guessing — you write what you remember happened, not what actually happened across the team. Our daily-management OS feeds the founder review with real Plan → Fact data from every team, so the 30-minute review is on signal, not on memory. See how the 7-day diagnostic surfaces your current Plan-Fact gap at https://aiadvisoryboard.me/?lang=en.

Manager scan (2-minute digest example)

Plan (last week's top 3):

• Tier-2 pricing finalized with CFO
• Hire #4 (Head of Customer Success) closed
• Q3 board memo first draft

Fact:

• Tier-2 pricing finalized
• Head of CS — verbal yes, contract Monday
• Board memo: only outline complete

Gap (the founder writes):

• Board memo slipped because pricing took two extra hours of CFO time — recurring "founder bottleneck" pattern
• Surprises this week: 4 (3 internal, 1 external) — the internal ratio is climbing, third week in a row

The Manager scan version exists for founders who later scale this to leadership team review. Same structure, same 30-minute timer.

Micro-case (what changes after 7-14 days)

A 90-person logistics SMB founder ran the 30-minute review on himself for four weeks. Week one took 55 minutes — too much detail, too many surprises documented as anecdotes. Week two he tightened the surprise count to five maximum. By week three the internal/external surprise ratio was 7:2, which made the pattern unignorable: most of his weekly surprises came from inside the company. He spent 45 minutes Friday afternoon mapping the internal-surprise causes; six of them traced to the same dispatcher escalation flow. One process change in week four cut the internal surprise count roughly in half, and by week six the review had become a planning tool instead of a debrief.

Note on this case: This example is illustrative — based on typical patterns we observe with companies of 30-500 employees, not a single named client. Specific numbers are rounded approximations of common ranges, not guarantees.

Tool tip (AIAdvisoryBoard.me): The compounding power of the weekly review is the four-week rolling read — when you can scan four artifacts in a row and see whether the internal-surprise ratio is improving or worsening. Doing this manually is fine for the founder; doing it across the leadership team is where it breaks. Plan → Fact → Gap operationalized across the company gives leadership the rolling read at every layer. Run the 7-day diagnostic at https://aiadvisoryboard.me/?lang=en.

FAQ

Can I do this in less than 30 minutes? Once it's stabilized, yes — around week six many founders compress to 20. Don't compress on week one. The discipline forms by completing all five questions, not by speed.

What if I missed a week? Don't backfill. Open the most recent artifact, run the current week's review, note "skipped W of [date]" in the surprises section. The cost of missing one week is small; the cost of breaking the artifact chain by faking entries is large.

Should I share the artifact with my leadership team? Only after eight to ten weeks of doing it for yourself. The personal version is honest. The published version becomes managed performance. Run it solo first.

How is this different from a journaling habit? Journaling is open-ended; this is structured around five fixed questions whose outputs feed each other. The structure is what makes it a management tool, not a reflection tool.

What if my week is too chaotic to review? Then your review's first finding is "the week was unstructured" — and Question 5's top priority next week is to restore one operating cadence. The chaotic-week excuse is itself the diagnosis.

Conclusion

The founder weekly review fails when it expands. It works when it's bounded by a timer, structured by five questions, and produces an artifact you re-read the next week. The artifact is the point. Without it, you're doing 52 disconnected debriefs instead of one continuous management practice.

Block 30 minutes this Sunday. Five questions. One page. Re-read it next week.

If you want a system that surfaces the Plan → Fact → Gap automatically — every day, so your weekly review starts on signal rather than memory — see how the 7-day diagnostic works at https://aiadvisoryboard.me/?lang=en.