It is normal for GA4, the ad platform and the CRM to show different numbers for the same campaign. Each measures differently. Instead of hunting for the right number, it pays to understand why they diverge and decide on the basis of a coherent set of sources.
The essentials in a definition
GA4 and ad platforms show different numbers because they use different attribution models, windows and consent rules. Reconciling does not mean equalising, but understanding the difference and choosing the right source for each decision. The key-event definition and platform conversion measurement explain each one’s criteria.
Why the numbers differ
Each system uses its own rules: attribution windows, models, handling of consent, deduplication, time zone, definition of conversion. The ad platform credits itself. Analytics may credit other channels. The CRM counts what was actually sold.
Even when everything is set up correctly, it is rare for the numbers to match. What matters is knowing what kind of difference it is.
A reconciliation sheet
For one month, list the numbers from each source: clicks, sessions, conversions, sales. Calculate the ratios between them and watch their stability over time. A stable ratio is predictable. A ratio that changes suddenly reveals a problem.
Note changes in configuration, site and consent that explain variations.
Reconciling numbers from several sources
- List numbers from each source for the same period
- Calculate the ratios between them
- Look for sudden changes and explain them
- Decide with the source closest to the business
Which source to use for what
For deciding bids and budgets in the short term, use the ad platform (it is what the algorithm sees). To compare channels, use analytics and tests. To assess the business, use the CRM and invoicing.
If the three tell different stories, favour the one closest to the money: real sales and margin.
The aim is not to make numbers match; it is to understand why they differ.
When to investigate
When ratios change without explanation, when platform conversions diverge widely from the CRM, or when a source suddenly disappears. Check for duplicate events, failing tags and site changes.
Avoid spending weeks trying to make numbers equal. A stable, understood difference is acceptable.
Decision table
| Source | Good for | Limit |
|---|---|---|
| Ad platform | Optimisation and campaign reporting | Attributes to its own campaigns |
| GA4 | Site behaviour and channel comparison | Depends on consent and model |
| Sales system or CRM | Real results and margin | May lack the customer’s source |
| Source survey | Qualitative clues | Subjective |
Common mistakes to avoid
- Expecting identical numbers. The sources answer different questions.
- Choosing the most favourable source. Define the reference source in advance.
- Ignoring configuration changes. Changes to events or model explain sudden jumps.
How to measure the result
Record monthly the ratio between GA4, platform and sales-system conversions; define alert thresholds for when the ratio departs from normal.
Limits and exceptions
Reconciling sources has limits: some differences are irreducible (consent, blockers, models). Chasing perfect equality wastes time without value. Define an acceptable ratio, monitor and investigate only when it departs from normal. Concentrate efforts on decisions where the difference changes the choice, and accept some residual uncertainty.
Illustrative example
Illustrative example: a company finds GA4 shows 70% of the conversions the ad platform records. Instead of trying to equalise, it calculates the typical ratio and tracks its stability, using the sales system as the final reference. When the ratio suddenly drops to 40%, it investigates and finds a consent change reduced tracking. The alert was possible because the ratio was recorded.
Frequently asked questions
Why does GA4 show fewer conversions than the platform?
Because of different attribution models, windows and consent handling, among other factors.
Which number should I use?
It depends on the decision: the platform for bids, analytics for comparison, the CRM for assessing the business.
Is a 20% gap serious?
Not necessarily, if it is stable and understood.
Which source should I use to decide budget?
The sales system and margin as reference, platforms to optimise and GA4 to analyse behaviour.
Why does GA4 show fewer conversions?
Because of consent, attribution model, window and event definition; a stable difference is normal.
Reconciling GA4 and platforms means knowing the typical difference, choosing a reference source per decision and investigating when the ratio changes suddenly.
Sources and recommended reading
- Analytics Help: GA4 recommended events
- Google Ads Help: conversion measurement
- Analytics Help: key events
Last editorial review: October 2026. Informational content; always confirm current conditions in the official sources before deciding.