Agenda (60 minutes)
- Results vs thresholds (15 min): companies, meetings, pipeline, share of answers.
- What worked (10 min): the top five buyer questions by meetings.
- What didn't (10 min): ad groups and pages to stop.
- Platform changes (10 min): targeting, bidding, policy and new AI ad platforms.
- Next quarter (15 min): budget, questions, pages and one experiment.
Platform changes to check
- The current ad policy version and any category changes.
- New targeting or bidding options.
- New markets for ChatGPT ads and AI Overview ads.
- Changes in how AI assistants cite sources.
The output
A one-page plan: budget by campaign, ten buyer questions to focus on, three pages to build or update, and the thresholds for the next review.
Pre-read for the meeting
- The last 13 weekly reports.
- The quarter's scorecard against thresholds.
- Sales feedback summary.
- A list of platform changes during the quarter.
Questions
Who should attend?
Marketing lead, sales lead, whoever runs campaigns and whoever owns the budget.
What's the one experiment for?
To keep learning: a new question type, a new country or a new landing page format.
Should AI visibility be in the same review?
Yes. Ads and visibility share the same questions and the same goal.
A quarter in review (fictional)
- Results: 340 companies identified, 118 in profile, 21 meetings, 8 opportunities; thresholds met.
- What worked: decision questions on desktop; the comparison page converted best.
- What didn't: problem-stage questions on mobile; paused.
- Platform changes: new bidding option available; ad policy unchanged for our category.
- Next quarter: +20% budget on the top two ad groups, a UK campaign, two new pages, and one experiment with conversion bidding.
Mistakes to avoid in reviews
- Reopening decisions already made without new data.
- Judging the channel on the worst week.
- Leaving without named owners for each action.
The review is only useful if next quarter looks different because of it.
Start here: AI ads for B2B. More articles in Measurement and reporting.