Answers

Why do my ad platforms and analytics disagree?

Your ad platforms and analytics disagree because each one counts a different thing, in a different way, over a different stretch of time. Meta, Google Ads, and GA4 are not three attempts at the same number. They are three scorekeepers with separate rulebooks, and none of them is lying to you.

So the real question is not which platform is right. It is which number you should trust for a given decision, and how far apart the others are allowed to drift before something is genuinely broken.

The four reasons the numbers never line up

Four forces drive the whole mess. Understand them once and the chaos turns into something you can predict and explain to anyone who asks.

Each platform attributes with its own model

Attribution is just the rule for who gets credit. Meta credits clicks and often views inside its window. Google Ads credits clicks, frequently with a data-driven model that spreads credit across touches. GA4 applies its own attribution and ties conversions to the session it believes drove them. Your CRM credits whoever actually closed the deal. Four rulebooks, one buyer, four different places the conversion can land.

Each platform uses its own attribution window

A window is how long after an ad a conversion still counts. The platforms use different defaults, and they let you change them. A sale that happens several days after the click gets claimed by a tool with a long window and quietly dropped by a tool with a short one. Same event, counted or ignored depending on whose clock you read.

Every platform grades its own homework

This is the big one. Meta and Google both want to prove they drove the result, so both will claim the same conversion. A buyer who saw a Meta ad, later clicked a Google ad, then converted, registers as a win in both accounts. Add the platform totals together and you double count every overlap. That is why the sum of your ad platforms almost always towers over what your CRM actually recorded.

Tracking loss and modeling fill the gaps differently

Cookie limits, iOS App Tracking Transparency, ad blockers, declined consent, and plain network failures mean some conversions never reach a given tool. Each platform patches its own hole with an estimate, called modeled conversions. GA4 models too. Your CRM does not model. It only holds what was truly recorded. So a modeled platform number will structurally sit above a hard-counted system of record, even when nothing is wrong.

Which number should you actually trust?

It depends entirely on the decision in front of you. There is no single right number because there is no single question. Match the tool to the job and the disagreement stops reading as a problem.

• For in-platform choices, which ad to scale, which audience to cut, when to pause, trust that platform's own numbers. Its algorithm optimizes toward its own signal, so that is the signal you steer with.
• For how much revenue marketing actually produced, trust the system closest to the money: your CRM, booking tool, or point of sale, tied back to campaigns with UTMs.
• For on-site behavior and the paths people take, trust GA4, while remembering it undercounts whenever consent is declined.

And never add conversions across platforms to build a grand total. The overlap gets counted twice and the total becomes fiction.

How to reconcile them instead of chasing a match

You are not trying to make three tools print an identical number. That will never happen and it is not the goal. You are trying to make them agree on direction and to explain every gap with a reason you can name.

• Define one conversion the same way in every tool: same event, same trigger, same page.
• Tag every campaign with consistent UTMs so GA4 and your CRM can see the source each platform is claiming.
• Name one source of truth for revenue decisions and write it down, so the choice is not relitigated in every monthly meeting.
• Use server-side tracking and a proper consent setup so fewer conversions are lost before modeling has to guess at them.
• Compare trend lines, not raw totals. If all three move the same direction week over week, your measurement is healthy even while the counts differ.

When a gap is normal and when it is a bug

A stable, persistent gap between a platform and your CRM is expected. It is the price of different models and modeling, and it does not need fixing. What should worry you is a gap that suddenly widens, a platform reporting conversions your CRM has no record of, or GA4 showing a traffic source that contradicts your UTMs. Those are symptoms of a broken tag, a misfiring pixel, or a tracking template that dropped its parameters. That is a real measurement problem, not an attribution quirk to shrug off. When we build and instrument a site, this reconciliation is part of the setup rather than an afterthought.