A data discrepancy is a difference between the figures two tools report for what looks like the same thing. For example, GA4 might show 40 purchases last month, Meta Ads Manager 55 and your shop platform 48. Some discrepancy between platforms is normal and expected; the skill is telling a normal gap from a broken setup.
How data discrepancies happen
Each platform counts in its own way, for its own purposes. The usual causes are:
- Attribution rules. At the time of writing (October 2026), Meta’s default attribution window credits a sale to an ad if the buyer clicked it within seven days or viewed it within one day. GA4 shares credit across channels using its own model, and Google Ads uses yet another view. Each platform tends to claim the sale for itself, so the totals add up to more than the number of sales you actually made.
- Dates. Google Ads reports a conversion against the day of the ad click, while GA4 reports it on the day it happened. A sale on Tuesday from a click on Saturday can land in different weeks.
- Consent and blocking. Visitors who decline cookies, use an ad blocker or browse with strict privacy settings may be missing from GA4 or appear only in modelled figures. Because analytics in the UK normally waits for consent, this consent-driven data loss is a bigger factor here than in markets with no banner requirement.
- Modelling. Ad platforms fill gaps with estimated conversions that your shop or CRM will never show.
- Payment journeys. When a customer pays through PayPal or completes a bank’s Strong Customer Authentication check, GA4 can credit the sale to the payment provider unless it is listed as an unwanted referral.
- Settings. A GA4 property and an ad account set to a different currency or time zone will disagree on daily totals and revenue.
- Tracking faults. Duplicate tags, a missing tag on one checkout path, or a thank-you page that fires again on refresh.
Why it matters
Discrepancies lead to two expensive mistakes. The first is believing the most flattering number. If Meta says a campaign returned £5 for every £1 spent and your bank account disagrees, scaling the budget on Meta’s figure loses money. The second is panicking over a gap that has always existed and rebuilding tracking that was working fine.
The useful question is whether the gap is stable. If GA4 has consistently recorded about four in every five of the orders in your shop platform for six months, that gap is a known feature of your setup. If it suddenly drops to half, something has broken, often after a site update, a new consent banner or a change to the checkout.
Common mistakes
- Expecting every platform to match exactly. They never will, and chasing a perfect match wastes time.
- Adding up platform-reported conversions from Google and Meta and treating the total as real sales.
- Comparing different date ranges, time zones or definitions, such as GA4 purchases against shop orders that include cancellations and refunds.
- Investigating without a source of truth. Your shop, booking system or CRM records what actually happened; analytics and ad platforms estimate where it came from.
- Ignoring a sudden change because “the numbers never match anyway”.
How to act on it
Choose a source of truth for each outcome: the ecommerce platform for orders, the CRM or inbox for enquiries, the phone system for calls. Each month, put that figure next to GA4 and each ad platform for the same dates and note the ratio. A simple table is enough. After a few months you will know your normal range, and anything outside it is worth investigating.
When the gap moves, check the obvious first: was a tag removed, the consent banner changed, a checkout step added or a payment provider switched? Then place a real test order or enquiry and follow it through each platform. I go through the most common version of this problem in why GA4 and Facebook conversion numbers do not match. Reconciling these figures is a standing part of performance marketing, because every budget decision depends on knowing which number to trust.