Skip to content
ROIMOBI
← All writing

Weekly Marketing Review: A 30-Minute Budget Agenda

By Vlad Yasinski·

A weekly marketing review that fits in 30 minutes needs a fixed agenda, a one-page pre-read and budget rules agreed before the meeting starts. Run the topics in this order: exceptions, channel payback and budget moves, tests and creative, market, then actions with owners. The meeting is done when each budget, test and kill call has a name and a date next to it.

Reviews fail when they report instead of decide. This 30-minute agenda forces a budget call, a test call and a kill call each week.

Layer 6 of the loop: DecideUser acquisition auditRequest an audit
Nov 2026·6 minOperating CadenceDecisionsGrowth

Key takeaways

  • A weekly marketing review works when it ends with three calls: where budget moves, which tests continue, and which tests stop.
  • A 30-minute agenda can split as 5 minutes on exceptions, 10 on channel payback and budget, 7 on tests and creative, 5 on market, and 3 on actions with owners.
  • Budget rules written before the meeting, such as raising spend while payback sits inside target and cutting after two misses, take the debate out of the room.
  • A one-page pre-read sent a day early and an action log with owners keep the meeting short and the follow-through visible.

Purpose and attendees

The purpose of the review is to decide. Reporting numbers belongs in the pre-read, so the meeting time goes to choices: move budget, continue or stop a test, assign a fix. A topic that produces no decision does not earn a slot on the agenda.

Keep the room small and cover three roles:

  • Budget owner: has the authority to raise, hold or cut spend in the meeting.
  • Creative owner: speaks for running tests, production capacity and the next brief.
  • Data owner: vouches for the numbers and explains any tracking gap or definition change.

On a small team one person can hold two roles, but all three need coverage. Other people read the notes afterward. Send the pre-read a day before, so attendees arrive having read it and the first minute goes to decisions instead of walking through tables. I owned budgeting, forecasting and P&L reporting on CAC, revenue, ROAS and payback, and the pre-read below tracks those same measures.

The pre-read

The pre-read is one page with the same layout each week, so eyes land on the same places and changes stand out. It carries five blocks:

  • Spend for the week and against plan, by channel.
  • New payers, counted from actual payments.
  • CAC by channel.
  • Payback by channel against the target.
  • Exceptions: anything outside the ranges you set.

Anomalies come from alerts, not from someone scanning charts on the morning of the meeting. Set thresholds for spend pacing, CAC and conversion rate by source and placement, and put the triggered alerts at the top of the page with one line each. A conversion rate that collapses for a single source or placement is often an early sign of weak or fraudulent traffic, and the alert list is where you catch it.

Close the page with a competitor digest of three lines: new angles, offers or pricing changes seen on public ads and pages. Three lines force a choice about what matters. Keep the digest to public data and respect platform terms.

The 30-minute agenda

Assign a timekeeper. If a topic runs over, park it in the action log with an owner instead of extending the meeting. The five slots run 5, 10, 7, 5 and 3 minutes, and 5 + 10 + 7 + 5 + 3 = 30. In order, they are:

  • Minutes 0 to 5 (5 minutes), exceptions: read last week's actions, then go through the alerts and decide for each whether it needs action this week or is noise.
  • Minutes 5 to 15 (10 minutes), channel payback and budget moves: read payback by channel against target, apply the decision rules and record each change as an amount.
  • Minutes 15 to 22 (7 minutes), tests and creative: review running tests, make the extend or kill call, and agree what goes into the next brief.
  • Minutes 22 to 27 (5 minutes), market: the competitor digest, plus any platform or seasonal change that affects spend.
  • Minutes 27 to 30 (3 minutes), actions and owners: read back each action with a name and a due date.

Budget gets the largest slice because money moves there. Exceptions come first because a tracking break invalidates the budget discussion that follows it.

Budget decision rules

Write the rules before the meeting so the room applies them instead of arguing about them. Three rules cover most weeks:

  • Raise spend on a channel when payback is inside target and has held for several weeks in a row.
  • Hold when payback is near the target or the data volume is too thin to trust.
  • Cut when payback exceeds target in two consecutive reviews. A single miss goes on watch.

Illustrative example with made-up numbers. Say your target payback is 60 days and a channel pays back in 48 days, steady for several weeks. It spends $10,000 a week. The rule allows a 20% raise: $10,000 x 0.20 = $2,000, so the new weekly budget is $12,000.

Check the headroom: 60 - 48 = 12 days. More spend usually pushes CAC up, which lengthens payback. As a worst-case assumption, say payback lengthens in the same proportion as the raise: 48 x 1.20 = 57.6 days, still under 60. Next week, if payback reads 58 days, hold the budget at $12,000. If it reads above 60, put the channel on watch, and a second miss triggers the cut.

Test and kill log

Keep one log for all tests, and fill in four fields before a test launches:

  • Hypothesis in one sentence, such as "a price shown in the first frame raises checkout starts".
  • Primary metric and the threshold that counts as a win.
  • Start date, end date and budget cap.
  • Owner.

The end date matters most. A test without one drifts, keeps spending and ends with an opinion instead of a result. At the review, each test due that week gets a call: extend with a reason, close as a win, or kill.

Code the kill reasons from a short list, for example no signal by end date, CAC above limit, payback too long, tracking fault, and creative fatigue. After a few weeks the codes show which reasons repeat. Many tracking faults point to a data problem, while many fatigue kills point to a production-capacity problem.

Winners go into the next brief. The creative owner takes the winning element, whether a hook, an offer or a format, into the next production brief, and the log records which brief received it.

Keeping it honest

Use the same definitions each week. Payback, new payer and CAC should be counted the same way each time. If a definition has to change, note it on the pre-read and restate the earlier weeks, because comparisons across weeks break otherwise.

Review last week's actions before anything new. Read each one aloud as done, open or dropped. This takes about a minute inside the exceptions slot, and it tells the room whether the previous meeting produced anything.

Skip charts with no decision attached. If a chart cannot change a budget, a test or a fix, it leaves the pre-read. A shorter page gets read, and a page that gets read makes the meeting shorter.

Next step: run the first review this week

  • Set a target payback for each channel and write the three budget rules next to it.
  • Build the one-page pre-read from the last four weeks of payments, spend and CAC.
  • Set alert thresholds for spend pacing, CAC and conversion rate by source.
  • Book a 30-minute slot at a fixed time with the three roles, and send the pre-read the day before.

After four reviews, count the decisions the log recorded against the meetings held. Remove any agenda item that produced none.

Want a review cadence backed by one revenue-based dashboard?

Book an audit