The real cost of "we'll handle tracking later"

Tracking is the most boring part of paid media. It is also the only part that determines whether everything else works. Every onboarding I do starts with the same conversation. The client wants to talk about the new campaigns. The new ad creative. The new landing pages. The geographic expansion. I want to talk about whether GA4 and Google Ads agree on what a conversion is. They never want to talk about that. Until three months later when nothing has scaled the way it should and we figure out together that the conversion data has been wrong the entire time. This is the most common pattern in the industry. Tracking gets pushed to the end of every project. The thinking is, let us get the campaigns live, see what happens, deal with measurement when we have time. The result is that the campaigns are live and optimising toward signals that do not match reality. Here is what "we will handle tracking later" actually costs. On Performance Max specifically, every campaign you run is using Google's machine learning to find conversions. If your conversion data is wrong, the machine is learning the wrong lessons. It is finding more of the wrong people. It is bidding higher for traffic that does not actually convert and lower for traffic that does. The longer you let this run, the more entrenched the wrong learning becomes. PMax does not unlearn. It just keeps reinforcing whatever pattern you fed it on day one. On Search campaigns the cost is different but real. If your conversion tracking is double counting, your reported ROAS is inflated. You think the campaign is working. You scale spend. Suddenly the inflated ROAS goes from 4x to 2x as the duplicates dilute, and you cannot figure out why scaling broke the campaign. Spoiler: it never worked at the original 4x. The numbers were lying. On lead generation accounts the cost compounds further. Marketing optimises toward form fills, not toward qualified leads. Sales burns through lists of leads who never actually wanted the product, just clicked the button by accident. Sales then tells you the leads are bad. You blame Google. Everyone is unhappy. The actual root cause is that the tracking was set up to count form submits as conversions when it should have been counting qualified pipeline. For ecommerce, the most expensive version is GA4 reporting different revenue from Shopify. I see this on almost every audit. GA4 says the store made $80k. Shopify says it made $120k. Where did the $40k go? Usually a combination of consent mode blocking events, duplicate purchase events causing deduplication problems, time zone mismatches, and refund handling differences. None of these are catastrophic on their own. Together they make every decision you make based on GA4 data slightly wrong. Slight wrong, multiplied across a year of campaign decisions, is a lot of wasted spend. The fix is not glamorous. It is hours of work checking that every event fires, every conversion has a unique transaction ID, every attribution window matches across platforms, every refund and cancellation flows through correctly. Server side tracking helps for ecommerce but is not always necessary. Sometimes a clean GTM setup with consent mode handles 90 percent of the problem. What I tell every new client during onboarding: before we touch any campaign, I am auditing the tracking. If it is broken, that is the first thing we fix. The campaigns can wait. Optimising broken data is worse than not optimising at all, because every change you make is locking in the wrong learning. If you are an in house marketer reading this and you have never seriously audited your tracking, do that this week. Pull the GA4 vs Shopify revenue comparison. Pull the Google Ads conversions vs GA4 key events. If there is a gap of more than 10 percent, you have a tracking problem. Fix it before you make another budget decision. If you are an agency reading this and you skip tracking audits during onboarding because the client is in a hurry to launch, you are setting up every campaign decision for the next year on a broken foundation. The hurry to launch is fake urgency. The tracking audit is the actual urgent thing.