- 1 List the events each platform optimises on
- 2 Walk the funnel with tag inspection on
- 3 Check each event fires once, with values
- 4 Test consent accepted and rejected
- 5 Fix, re-test, then raise budget
Ad platforms now decide who sees your ads based on the conversions you send back to them. If those signals are wrong, the algorithm optimises towards the wrong people, and more budget only makes the mistake bigger. A tracking audit before every budget increase is the cheapest insurance in paid acquisition.
Why it works
Smart bidding and automated audiences learn from your conversion events. A purchase that fires twice tells the platform each sale was worth double. A lead event that fires on page load, not on form submit, rewards clicks instead of leads. A missing value parameter makes every order look identical, so value-based bidding has nothing to work with.
These errors are silent. Dashboards still fill up with numbers, which is exactly why they survive for months. A structured walk through the funnel surfaces them in under an hour, before they compound across a larger budget.
How to do it
- List the events that drive spend. For each ad platform, write down the conversion events it optimises on (for example
purchase,lead,sign_up) and the parameters each one needs: value, currency, transaction or event ID. - Walk the funnel like a customer. Open the site with a tag inspection tool running. Land from a tagged ad link, browse, add to cart or start the form, and complete a test order or test lead.
- Check each event fires exactly once. Look for the same event firing twice on one action. Duplicates usually come from a hard-coded pixel plus a tag manager tag, or a thank-you page that fires again on refresh.
- Check the parameters. Value should match the order total, currency should be set, and IDs should be present so platforms can deduplicate browser and server events.
- Test both consent states. Reject cookies and repeat the journey, then accept and repeat. Tags should respect the choice, and if you use consent mode, the right consent signals should be sent in both cases.
- Confirm in each platform’s own tool. Use GA4 DebugView and each ad platform’s test-events view to confirm what the platform received, which can differ from what the browser sent.
- Write down what you found and fixed. Keep a one-page log with the date, the issue and the fix. It becomes your pre-flight checklist for the next budget increase.
Example
Say a D2C brand plans to double its paid social budget. Before raising it, the team walks through a test checkout and finds purchase firing twice, once from the platform’s pixel code pasted into the theme years ago and once from the tag manager. Reported return on ad spend had been roughly twice the real figure. They remove the hard-coded pixel, confirm one event per order with the correct value, and only then scale. The higher budget now trains bidding on real sales.
A B2B SaaS company does the same audit and finds its lead event attached to the demo page view instead of the form submission. The fix cuts reported leads but makes cost per lead honest, and the campaigns are re-optimised on people who actually requested a demo.
Metrics to track
- Events per conversion. Platform-reported conversions divided by real orders or leads in your back end. It should sit close to 1.
- Share of conversion events with value and currency set. Aim for all of them on purchase events.
- Gap between platform-reported and back-end revenue. Some gap is normal because of attribution windows, but a sudden jump usually means a tracking change.
- Days since last audit. Keep it under a month, and reset it after any release that touches checkout or forms.
When it backfires
- Auditing once and forgetting. Tracking breaks without warning after theme updates, new apps and tag manager edits. A single audit only proves it worked on that day.
- Fixing in the browser only. If you also send server-side conversions, a change on one side without the matching event ID on the other can create the duplicates you just removed.
- Testing with your own ad blocker on. Blockers hide tags and make healthy setups look broken. Test in a clean browser profile.
- Changing tracking mid-test. Fixing conversion events changes the numbers the platform learns from. Expect a short re-learning period and avoid judging campaigns in the days right after the fix.
Growth Marketing Academy
Learn the framework behind this tactic
These lessons explain the math and the mental model, so you can adapt the tactic instead of copying it.
- Track 2 · Acquisition Paid Acquisition Architecture and Channel-Model Fit Structure paid accounts, run a creative testing cadence, read CAC, blended CAC, MER and ROAS correctly, and pick channels your price point can afford.
- Track 2 · Acquisition Growth Loops: Build Acquisition That Compounds Funnels describe how users leak. Loops describe how users create more users. Learn to map, quantify and choose the loop that fits your product.
Questions people ask
How often should I audit my tracking?
Before any meaningful budget increase, after every site release that touches checkout, signup or the tag manager container, and at least once a month as a routine check.
What is the most common tracking problem?
Duplicate conversions. The same purchase or lead fires twice, often once from a hard-coded pixel and once from the tag manager, which makes campaigns look better than they are and teaches bidding algorithms the wrong lesson.
Do I need developer help to audit tracking?
Not to find problems. You can walk the funnel yourself with a tag inspection tool and a test order or test lead. Fixing what you find may need someone with access to the tag manager or site code.