How to run a 30 minute weekly marketing review

1. Make it a decision meeting, not a reporting ceremony

A weekly marketing review should answer three questions: what needs attention, what should change, and who will act? Reading every dashboard aloud consumes the time needed for judgement. Circulate routine updates beforehand and reserve the meeting for exceptions, competing priorities and approvals.

Invite the people who can interpret the evidence, commit delivery capacity and authorise changes. One person may cover several roles. Name a chair and a decision recorder; neither role requires a new job title. If an approver cannot attend, record the approval dependency rather than treating silence as permission.

The following examples are hypothetical. For a local plumbing business, the useful question is not whether clicks increased. It is whether qualified emergency-call enquiries justify extending an evening campaign when the roster can actually service those calls.

• Agree one primary commercial outcome, such as qualified enquiries or fulfilled orders.
• Set a practical limit of three new commitments, adjusted to available capacity.
• Move detailed troubleshooting into a separate session with a named owner.

2. Prepare one page before the timer starts

Use a repeatable pre-read: outcome against plan, spend against budget, important delivery changes, unresolved decisions and source-health warnings. Every metric needs its source, reporting window and definition. Compare equivalent weekdays and flag promotions, stock shortages or public holidays that weaken comparability.

Keep GA4 key events and Google Ads conversions distinctly labelled. Google explains that they differ; a key event can be used to create a Google Ads conversion. Do not assume similarly named columns are interchangeable or add them together as unique outcomes.

For an ecommerce team, show paid spend, orders from the chosen business record and refund context separately. If Monday's figures are incomplete, mark them provisional. A polished but stale dashboard is less useful than a plain table that admits its limits.

• Include absolute totals beside percentage changes, especially with small volumes.
• Show the previous meeting's commitments and their status.
• Ask owners to submit decision requests, not slide decks.

3. Allocate all 30 minutes before you begin

Display the agenda and use a visible timer. The chair should interrupt repeated context politely: 'What decision do you need, and what evidence would change it?' Save unresolved technical questions in a follow-up list with deadlines, not an unowned parking lot.

For a hospitality venue, a missing booking feed may make the campaign discussion inconclusive. Use the source-health slot to establish that limitation, then decide who will check bookings manually. Do not spend the entire meeting trying to repair the connector.

• Minutes 0–3: confirm the objective, source freshness and urgent risks.
• Minutes 3–10: review outcome and budget exceptions.
• Minutes 10–17: inspect completed work and active tests.
• Minutes 17–24: rank opportunities and resolve delivery blockers.
• Minutes 24–28: assign actions, capacity and approvals.
• Minutes 28–30: read back decisions and review dates.

4. Separate business problems from measurement problems

Use an exception rule agreed in advance: discuss a deviation when it could change a commercial decision, threaten a budget limit or expose unreliable measurement. There is no universal acceptable CPA. The relevant range depends on lead quality, margin, fulfilment costs and the business's financial constraints.

Suppose a professional services campaign has fewer form submissions but more sales-qualified consultations. Before cutting spend, inspect lead qualification and reporting delay. Conversely, cheaper leads may be worse if sales rejects most of them. Ask for counts and definitions before explanations.

Search deserves the same restraint. Google's Search Console documentation describes aggregation and average position, so avoid treating a site-wide position change as a fixed ranking movement. A retailer might inspect the affected queries and pages before rewriting a category page.

Label each exception as performance, measurement or unresolved. An unresolved issue warrants an investigation task, not a confident budget recommendation.

5. Review completed work without claiming it caused every change

For each completed action, record what shipped, when it shipped, the intended outcome and what changed afterwards. Then list plausible alternative explanations. Period comparisons describe movement; they do not isolate the effect of your work.

Suppose a software team changes its trial page while also launching a partner newsletter. More sign-ups the following week do not establish that the page edit caused the increase. Keep the observation, but mark attribution uncertain and decide whether a controlled test is worthwhile.

Google Ads experiments can compare an experiment with an original campaign. Where appropriate, use that structure rather than repeatedly editing the baseline. However, a weekly meeting is a review cadence, not a requirement to conclude every experiment within seven days.

• Check the original hypothesis and success measure before inspecting results.
• Record overlapping changes, tracking issues and conversion delays.
• Choose continue, investigate, adopt or stop, with a reason and next checkpoint.

6. Rank opportunities against real delivery capacity

Compare likely commercial impact, confidence, effort and urgency. Keep estimates visibly separate from measured results. A rough ranking can organise discussion, but spurious numerical precision should not override a broken checkout or a contractual deadline.

Suppose a retailer must choose between correcting misleading delivery information and commissioning a new video suite. The delivery correction may remove an immediate customer obstacle, while video offers uncertain upside and needs approvals. Make that tradeoff explicit rather than choosing whichever task has the most enthusiastic sponsor.

Check the bottleneck person's capacity, not just the team's total hours. Three apparently small tasks may all require the same designer. When accepting work, identify what will wait. A review that only adds priorities produces a longer queue, not a clearer plan.

• Accept: evidence and capacity justify action now.
• Park: record the trigger or date for reconsideration.
• Dismiss: retain the reason so the same suggestion is not debated next week.

7. Turn decisions into testable commitments

Write actions during the meeting. Each needs one accountable owner, a due date, a completion criterion, an approval route and a measurement checkpoint. Separate delivery completion from business success: publishing a revised page is observable immediately; its commercial effect may take longer to assess.

For an agency client, 'improve landing-page conversion' is not an executable task. A better commitment is: 'Account manager to obtain client approval for revised eligibility copy by Thursday; web owner to publish after approval and verify enquiry submission; analyst to review results after the agreed observation window.'

Read the commitments back in the final two minutes. Ask each owner to confirm capacity and surface dependencies. If approval remains outstanding, describe the work as awaiting approval, not ready to launch.

8. Measure efficiency without hiding preparation time

Track total person-time: attendance multiplied by meeting duration, plus preparation and follow-up. Suppose four people previously met for 60 minutes and spent 80 combined minutes preparing. That is 320 person-minutes. A 30-minute meeting with 100 minutes of preparation and follow-up totals 220, a hypothetical saving of 100 person-minutes.

That difference is not automatically a financial saving or proof of better marketing. Check whether attendance changed, reporting work moved elsewhere or decisions became rushed. Record time consistently across several comparable weeks.

Balance time with usefulness: actions completed by their due date, decisions reopened because evidence was missing, and time from blocker identification to resolution. A shorter meeting that causes repeated rework is not an efficiency improvement. These measures can reveal progress, but no workflow guarantees it.

9. Use tools to preserve evidence and human judgement

Its described workflow fits this review structure: DailyRise ranks evidence-backed opportunities; accepted opportunities become prepared RiseProjects tasks; RiseInsights brings together named sources, reporting windows and comparisons of predicted and measured impact.

Use those comparisons as review evidence, not causal proof. Your team still decides priorities, assigns capacity and approves action. See DailyRise, RiseProjects and RiseInsights through the links below, or explore the full workflow at /the-solution.

Connected evidence depends on provider permissions, availability and approvals. TikTok Ads and LinkedIn are approval pending; Microsoft Advertising is planned. Do not build a mandatory review step around an unavailable connection. Keep a documented manual fallback and identify its owner.