--- name: paid-media-review description: Review paid advertising (Google Ads, Meta, TikTok, LinkedIn, Snapchat) by what it produces in booked clients and revenue, not clicks, and recommend where to move budget. Use when someone shares ad account exports, CRM data or a spend summary and asks whether their ads are working, or how to split a budget. --- # Paid media review Ad platforms report their own results, and each tends to take credit generously. This review judges spend on what the business itself records: enquiries that became clients, and the revenue they brought. Where that data is missing, the first recommendation is to start collecting it. ## Data to collect Ask for, in this order of value: 1. Spend by platform and campaign for the period, from each ad account. 2. Leads from the CRM or lead sheet, with source, campaign, date, status (booked, lost, open) and value where known. 3. What each platform reports as conversions, and how each conversion is defined. 4. Targets: an acceptable cost per booked client, or a return on ad spend target. Use at least one full sales cycle of data. For businesses with long cycles, such as property or B2B, a single month will undercount bookings from recent spend; say so in the report. ## Analysis Compute these per platform and per campaign, and show the working: - **Cost per enquiry**: spend divided by enquiries. - **Booking rate**: booked divided by enquiries. - **Cost per booked client**: spend divided by booked clients. - **Return on ad spend**: booked revenue divided by spend, where values exist. - **Platform gap**: platform-reported conversions compared with CRM enquiries from the same source. A large gap points to tracking problems or duplicate counting. Then look for: - Campaigns with spend and enquiries but no bookings over a full cycle. - Brand search campaigns mixed with non-brand ones. Report them separately, because people searching for the brand name were likely to find it anyway. - Lead forms inside Meta or LinkedIn that bring many contacts but few replies. Compare their booking rate with website and WhatsApp leads. - Conversion actions that optimise for the wrong thing, such as page views or clicks. - Whether booked clients are sent back to the platforms (offline conversions in Google Ads, Conversions API for Meta). Without that, each platform optimises for enquiries regardless of quality. ## Recommendations - Move budget in steps, not all at once. Suggest a change size and a review date. - Protect campaigns that produce booked clients at or under target, even if their cost per enquiry looks high. - Cut or fix campaigns with no bookings after a full cycle, and say which fix to try first (audience, offer, landing page or follow-up). - If follow-up is slow, say so: a faster reply often improves results more than a bid change. - Separate learning budget from proven budget so new tests are judged fairly. ## Output 1. **Headline**: in three sentences, what the spend produced and the most important change. 2. **Table** by platform and campaign: spend, enquiries, booked, cost per booked client, revenue, return on ad spend. 3. **What is working**, **what is not**, and **what we cannot tell yet** because of missing data. 4. **Budget plan** for the next period with amounts or percentages and a review date. 5. **Tracking fixes** needed to make the next review more accurate. State the currency, the date range and every assumption. Do not estimate figures the data does not contain. Do not change live campaigns or budgets yourself; recommend, and let the account owner apply changes.