The weekly founder review: what to continue, change or stop
Turn your weekly review into decisions: continue, change or stop. A worked fictional review and reusable evidence-to-action template.
A weekly founder review should end with decisions: what to continue, what to change and what to stop. Start with the goals, inspect the evidence and assign the next actions. A recap of everything that happened is useful only if it helps you choose what happens next.
AI can help assemble notes and prepare a first pass. The founder still needs to challenge the interpretation, resolve competing priorities and decide where authority or resources should change. The review is where the operating loop reconnects to strategy.
Prepare evidence before the meeting
For each priority, collect the intended outcome, the current owner, the last decision and the best available evidence. Include unresolved questions and exceptions. If a number is estimated or a status has not been checked, mark that clearly.
Keep the packet short enough to inspect. Link to supporting material rather than pasting every conversation into the review. The reader should be able to move from a claim to its source without having to reconstruct the entire week.
Then ask a more useful question than “Was everyone busy?” Ask whether the work has changed the position of the business in the way the goal intended.
A reusable weekly review
| Review field | Question to answer |
|---|---|
| Goal | What outcome were we trying to achieve? |
| Evidence | What actually changed, and how do we know? |
| Exceptions | What failed, remains unknown or crossed a boundary? |
| Decision | Continue, change or stop—and why? |
| Next action | What specific action follows from that decision? |
| Owner and date | Who owns it, and when will the evidence be checked? |
| Authority | Does the action fit the current permission, or require a new decision? |
Use one row per priority. If a row contains several unrelated decisions, split it. That makes ownership and follow-through easier to inspect.
A worked founder review
This is a fictional review of three small initiatives. The observations are illustrative, not a report of Ben's businesses or Walter performance.
Continue: the onboarding preparation trial
The team tested a draft onboarding pack against its acceptance checklist. Missing inputs remained labelled, and the manager could identify the owner for each next step. The decision is to continue with one further supervised example under the same draft-only boundary.
The owner is the operations manager. The next action is to choose the next authorised test case and record any new gaps. Continuing does not mean expanding permissions or declaring the process fully automated.
Change: the project-status briefing
The briefing included a provisional date as though it were confirmed. The decision is to change the source-status rule and rerun the fictional case that exposed the problem. Until then, the briefing remains internal and subject to review.
The owner is the project lead. The next check asks whether proposed, confirmed and disputed statements remain distinguishable. “Improve the prompt” would be too vague to track; this change names the actual failure.
Stop: the unused summary
A recurring summary is being produced, but nobody uses it to make a decision. The team cannot name the action it should support. The decision is to stop producing it while the intended user clarifies the need.
Stopping is not necessarily a technical failure. It can be the correct response when an output has no clear purpose. Preserve the useful source material and the reason for the decision rather than letting the job disappear without explanation.
Ask where the strategy itself needs attention
Not every exception should be fixed inside the existing plan. Sometimes the evidence challenges the priority, deadline or assumption underneath it. Leave space to ask whether the goal still matters and whether the proposed outcome is achievable with the current resources.
Distinguish a tactical adjustment from a strategic decision. Changing a summary format is one thing. Deciding to pursue a different customer segment is another. AI can help prepare alternatives, but a more consequential choice needs the appropriate people and evidence.
Close the review with an action record
Before finishing, read back the decisions, owners and dates. Make sure “continue” means a defined next step, “change” names the thing being changed and “stop” identifies the work that will cease. Record any follow-up that still needs approval.
At the next review, start with those decisions. Did the action occur? What did it teach you? If the same unresolved item keeps returning, identify the missing authority, input or capacity rather than copying it forward indefinitely.
Your weekly review does not need to be elaborate. It needs enough discipline to turn evidence into a decision and a decision into accountable action. That is how the loop stays connected to the direction you intended.
Explore Always Running: visit AlwaysRunning.ai to learn more.
