The short version
- GA4 and your CRM will never match exactly, and chasing an exact match is a waste of time.
- They measure different things. GA4 counts events in browsers. Your CRM counts records created by humans and systems.
- Aim for a stable, explained gap rather than a perfect one. Anything within a consistent range is usually fine.
- The CRM is the source of truth for revenue. GA4 is the source of truth for behaviour. Do not ask either to do the other’s job.
Someone opens GA4, sees ninety conversions, opens the CRM, sees sixty two enquiries, and asks a perfectly reasonable question: which one is wrong?
Usually neither. They are counting different things, in different places, at different moments, with different rules about what counts. Understanding the causes turns an alarming discrepancy into a known and boring one, which is what you want.
Nine reasons the numbers differ
1. Consent
If a visitor declines analytics cookies, GA4 either does not see them or sees a modelled estimate. Your CRM sees the enquiry regardless, because the form was submitted. This alone can account for a large share of the gap, and the share varies by how your consent banner is worded and positioned.
2. Ad blockers and privacy browsers
A meaningful slice of traffic blocks analytics outright. Those people can still fill in your form. Every one of them is a CRM record with no GA4 event behind it.
3. Duplicate submissions
Someone submits the form, the page is slow, they press it again. GA4 records two events. Your CRM either records two records or deduplicates them, depending on how it is configured. This pushes the numbers in the opposite direction to the first two causes, which is why the net gap can look small while both underlying errors are large.
4. Test and internal traffic
Your own team testing the form, a developer checking a fix, an agency running a QA pass. If internal traffic is not filtered from GA4 and test records are deleted from the CRM, they will disagree.
5. Spam
Bot submissions may be recorded by GA4 as conversions if they execute JavaScript, and blocked before reaching the CRM if you filter them. Or the reverse. Either way it is noise in the gap, and it is worth knowing which side of the wall your filtering sits on.
6. Timing and time zones
GA4 attributes an event to the session date in its configured time zone. Your CRM stamps the record when it was created, in its own time zone, possibly after a queue or an integration delay. At the end of a month, records slide between periods.
7. Attribution windows
GA4 credits a conversion to a session that may have started days earlier. Your CRM has no concept of a session at all. Comparing a channel report to a CRM lead source field is comparing two different models of causation.
8. Phone and email enquiries
Somebody rings you or emails directly. That is a CRM record with no website event behind it unless you have call tracking in place. For many businesses this is the single largest component of the gap.
9. Integration failures
The least comfortable cause. If the connection between form and CRM fails silently, enquiries exist in GA4 and nowhere else. This is worth ruling out first, because it is the only cause on this list that is actively costing you money right now.
Before investigating anything else, submit a real test enquiry and confirm it arrives in the CRM, in the inbox, and in GA4. If any of the three is missing, you have found something more urgent than a reporting discrepancy.
What an acceptable gap looks like
There is no universal correct figure, and anyone who quotes you one is inventing it. What matters is that the gap is stable and explained.
Track the ratio between the two counts every month. A gap that sits in a consistent range and moves gently is a healthy sign, because it means the causes are structural and understood. A gap that jumps sharply between months is a signal that something has broken: a consent banner change, a form change, a tag that stopped firing, a CRM automation someone added on a Friday afternoon.
In other words, monitor the derivative, not the value. The month the ratio changes is the month to investigate.
Numbers that will not reconcile? We audit the whole chain, from tag to form to CRM, and give you a written account of where each enquiry is and is not being counted. Most audits find at least one silent failure.
Which number do you report to the board?
Use each system for what it is genuinely good at.
- Revenue, pipeline, closing rates and cost per customer come from the CRM. It is the only system that knows what happened after the enquiry. Every commercial number should trace back to it.
- Behaviour, channel mix, page performance and campaign trends come from GA4. It is the only system that saw the journey. Use it to understand why, not how much.
- Cost per customer by channel needs both. That number requires the click identifier to survive the trip into the CRM, which is a separate piece of work described in our guide to tracking which marketing generates revenue.
Reporting one figure for enquiries and being clear about which system produced it removes most of the argument. Reporting two figures without explanation removes most of the trust.
A reconciliation you can run monthly
Half an hour a month, and it will catch problems long before anyone notices them in the revenue.
- Pull total form submissions from GA4 for the month.
- Pull total enquiry records created in the CRM for the same period, from web sources only.
- Subtract known non-web enquiries from the CRM count: phone, email, referral, walk-in.
- Subtract known test and spam records from both sides.
- Record the ratio. Add it to a running sheet.
- Investigate only when the ratio moves outside its usual range.
Six months of that sheet is worth more than any one-off audit, because it turns a vague suspicion that the numbers are wrong into a dated record of when they changed.
Common questions
Should I switch to server-side tagging to fix this?
Server-side tagging can reduce losses from blockers and improve data quality, but it does not remove the need for consent and it will not close the gap entirely. It is a worthwhile improvement, not a cure, and it adds infrastructure you then have to maintain.
Can I just use the CRM and ignore GA4?
You can report commercially from the CRM alone, and many businesses should. What you lose is the ability to see what happened before the enquiry, which is what tells you where to spend next.
Why did my numbers change when we updated the cookie banner?
Because consent rates changed. This is normal and expected. Note the date on your running sheet so future comparisons account for it, rather than treating it as a mystery.
Our CRM lead source field is mostly blank. Is that fixable?
Yes, and it is usually the highest-value fix available. A blank lead source means the enquiry arrived without its origin attached, which is a form and integration problem rather than a CRM problem. It is normally a small piece of integration work.
