You open Google Ads and it says your campaign got a certain number of clicks yesterday. You open Google Analytics and the visits from that campaign are noticeably fewer. Meta Ads Manager reports more purchases than your shop actually received. Then you look at the orders in your own system, and that number matches none of them.
It is tempting to assume something is broken, or that someone is inflating figures. Usually neither is true. Each tool counts different things, at different moments, by different rules. Once you know what those rules are, the gaps stop being alarming and start being useful.
A click is not a visit
An ad platform counts a click when someone taps the ad. Your analytics tool counts a session only when the page loads far enough for its tracking code to run, and only if that code is allowed to send data. A lot can happen in between.
- People leave before the page loads. On a slow page, or a weak mobile connection, someone may tap, wait, and hit back. The ad platform counted the click. Analytics never saw a visitor.
- Tracking is blocked. Ad blockers, privacy-focused browsers and some browser settings stop analytics scripts from loading at all.
- Consent was declined. If your site uses a cookie banner, visitors who refuse analytics cookies may not be recorded, or may be recorded only in a limited way, depending on how your consent setup works.
- Redirects lose the tracking. If the ad points to an address that redirects, the redirect can strip the tracking parameters, so the visit shows up as direct traffic instead of coming from the ad.
If the gap between clicks and sessions is large, a slow landing page is worth checking first. It costs you twice: the number looks wrong, and the people who gave up never saw your offer.
Each platform gives itself credit in its own way
This is the biggest reason conversion numbers never agree. A customer might see a Facebook ad on Monday, search your business name on Wednesday, click a Google ad, and buy on Friday after typing your address directly. Who gets the sale?
Each tool answers that differently, and each tends to answer in its own favour:
- Google Ads credits conversions to its own ads, using the attribution model and conversion window set on each conversion action.
- Meta can count a purchase if someone clicked an ad within its attribution window, and it can also count view-through conversions, where the person only saw the ad and never clicked it.
- GA4 looks at all your traffic sources together and divides the credit by its own attribution settings, so it may give part or all of the sale to a channel neither ad platform knows about.
So one real sale can show up as a conversion in Google Ads, another conversion in Meta, and a third credit in GA4 under a different channel. Add the ad platforms together and you will often count more sales than you made. That is not fraud. It is overlapping claims to the same customer.
Dates and time zones
Two smaller things cause a surprising amount of confusion.
First, time zones. Each account has its own time zone setting. If your Google Ads account, your Meta ad account, GA4 and your shop are not all on the same one, “yesterday” means a slightly different stretch of hours in each. Daily figures will drift even when the totals for a month look close.
Second, which date a conversion belongs to. Google Ads generally reports a conversion against the date of the ad click, while GA4 reports it on the day the conversion happened. If someone clicks on the last day of the month and buys a few days later, the two tools file that sale in different months.
Tags that fire twice, or not at all
Many mismatches come from the setup itself rather than from the platforms.
- Duplicate tags. The same conversion tag installed twice, once by a plugin and once by hand or through Google Tag Manager, will count each sale twice.
- Thank-you page reloads. If the conversion fires on a confirmation page, a customer who refreshes it or returns to it later can be counted again.
- Missing tags. A theme change or a new checkout plugin can remove the tag without anyone noticing, and conversions quietly drop.
- Payment pages off your site. If customers pay on a gateway’s page and do not come back to your site afterwards, the confirmation tag may never fire.
If you are not sure how analytics was put on your site in the first place, our guide to adding Google Analytics to WordPress explains the usual methods, which also helps you spot when two of them are running at once.
Modelled conversions
Because some visitors cannot be tracked, Google and Meta both fill part of the gap with modelling. They estimate conversions they could not observe directly, based on patterns from the ones they could. Google Ads uses modelled conversions in this way, and it ties in with consent mode for visitors who decline cookies.
Modelled conversions are a reasonable estimate, not a list of real orders. You cannot match them one by one against your sales, and they are one more reason the platforms will never agree exactly with your shop.
Bots
Not every click is a person. Google Ads filters clicks it judges invalid and does not charge for them, and GA4 automatically excludes traffic from known bots. Neither catches everything, and each filters in its own way, so a bot visit removed in one tool can still appear in another. A burst of very short visits from one campaign, with no enquiries at all, is worth a closer look.
What to actually do about it
Pick one source of truth for sales
Real money comes from real orders and real enquiries. Use your shop’s order list, your CRM or your enquiry inbox as the final word on how many sales you made. The ad platforms and analytics are there to tell you where those sales probably came from, not how many there were.
Compare trends, not exact numbers
The figures will never line up, so stop trying to make them. Instead, watch whether they move together. If Google Ads clicks rise and GA4 sessions from Google Ads rise too, things are working. If clicks rise and sessions fall, or the gap suddenly widens, something has changed and is worth investigating.
Check your tracking setup
Every so often, place a test order or send a test enquiry and confirm it appears once, and only once, in each tool. Check again after any theme change, checkout change or new plugin, because those are the moments tags get lost or doubled.
Tag your links with UTM parameters
UTM tags are short labels added to a link that tell analytics exactly which campaign sent the visit. Google Ads can tag links automatically when auto-tagging is switched on, but for Meta, email newsletters and other sources you usually add them yourself. Our free UTM builder makes the links for you, and keeping the names consistent means your reports group campaigns correctly instead of scattering them.
Make sure the landing page holds people
Since a slow or confusing page is where many clicks turn into nothing, it is worth running your ad landing pages through a website speed test now and then, especially on mobile.
If you are paying for ads and the page they land on is slow, unclear or not tracking properly, that is the part we can help with. Our Google Ads landing page design service builds pages that load quickly, match what the ad promises and have conversion tracking set up and tested, so the numbers you see come from tracking that has actually been checked.
Next in this series: Google Tag Manager Explained for Business Owners
