The Conversion Tracking Audit That Explains Why Your Best Campaigns Might Be a Mirage

I've walked into more accounts than I can count where one campaign looks like a machine and another looks like dead weight — and the tracking, not the targeting, is the reason. Nobody questions it because the numbers are right there in the dashboard. But dashboards report what they're told to report. If the tracking setup is double-counting, misattributing, or silently dropping offline conversions, you're not optimizing a business. You're optimizing a rounding error.

This is the audit almost no agency runs, because it doesn't touch a bid or a budget. It's not exciting. It doesn't produce a screenshot for a client call. But it's the layer everything else sits on top of, and if it's wrong, every optimization built on top of it is wrong too.

The most common inflation I find isn't fraud or manipulation — it's just duplicate tagging. A conversion action fires on a thank-you page and also gets imported from a CRM once the lead closes. Or Google Ads and Google Analytics are both pushed into the same optimization set through a shared tag. Each one, on its own, is reasonable. Together, they double the reported number for the same real-world event.

• Same conversion counted via both the on-site tag and a CRM/offline import
• GA4 conversions imported into Google Ads alongside a native Ads tag for the same action
• Multiple 'Purchase' tags firing on page refresh or back-button navigation
• Phone call conversions counted from both call tracking software and Google's own call extension

Run this check first: pull your conversion actions list in Google Ads and map every single one to the actual business event it represents. If two actions map to the same event, you have a duplication problem, and whichever campaign happens to touch that event twice is going to look artificially efficient.

The second failure point is attribution windows set on autopilot. Google Ads defaults skew generous, and most accounts never revisit them. If your sales cycle for a B2B lead is 45 days but your conversion action is set to a 30-day click window, you're either missing late conversions entirely or, more commonly, you've got a mismatch between what the platform is optimizing toward and what your CRM eventually shows as revenue. Ecommerce accounts have the opposite issue — a 90-day window on an impulse-purchase product blurs credit across campaigns that had nothing to do with the actual decision.

A campaign that looks brilliant on a 90-day view-through window and mediocre on a 7-day click window isn't two different campaigns — it's one campaign wearing two different masks.

Check every conversion action's attribution model and lookback window against your actual sales cycle, not the platform default. If you've read Lessons learned in automated bidding for PPC, you already know Smart Bidding is only as good as the signal you feed it — a mismatched window is feeding it noise and calling it a signal.

This is the one that actually costs money, not just accuracy. Top-of-funnel campaigns generate leads that convert to revenue weeks later, offline, in a CRM. If that offline conversion import is broken, delayed, or only partially mapped, the campaign that's actually driving real revenue looks like it's underperforming — while a campaign generating instant, low-value form fills looks like the star. Budget follows the mirage. The real revenue driver gets starved.

If you're running lead gen, this is worth pulling apart the same way I described in The Search Term Mining Framework That Catches Wasted Spend Before It Compounds — except instead of auditing where spend leaks out, you're auditing where credit leaks out of your reporting.