GA4 and ad platforms show different numbers for the same campaigns, and this causes distrust and arguments. The differences come from attribution models, time windows, consent and distinct definitions. Understanding them avoids decisions based on unfair comparisons.
The essentials in a definition
Attribution in GA4 and in ad platforms gives different numbers because it uses different models, windows, counting rules and data sources. Platforms tend to attribute to their own activity, GA4 applies its own rules to all channels. Understanding the differences avoids decisions based on wrong comparisons. Conversion measurement and consistent key events are the basis, and a dashboard (for example, Looker Studio) helps compare.
Why numbers differ
Platforms attribute conversions to their own campaigns, often with view-through attribution and long windows. GA4 measures from the site, attributes under another model and depends on consent and blockers. The CRM or sales system may have yet another count. Each source answers a different question.
None is absolute “truth”.
Models and windows
Last-click attribution, data-driven, with click and view windows: each choice changes the credit. GA4 and the platforms also differ in the date assigned to the conversion (date of interaction or of conversion). Record the definitions used in each report and compare under the same conditions.
Compare like with like.
Attribution in GA4 and on platforms: understanding the differences
- Understand that each source answers a question
- Record models and windows
- Use the CRM as the reference for results
- Decide by trends and margin
How to reconcile
Use the sales system or CRM as the reference for real results. Use the platforms for optimisation and GA4 for analysis of site behaviour. Estimate the typical ratio between sources and track its stability. If the difference changes suddenly, investigate tracking. Use incrementality tests to validate the effect.
The aim is not to equalise, it is to understand.
Numbers can only be compared when they answer the same question.
Practical decisions
Avoid cutting channels based on a single report. Observe trends, not only absolute values. Cross-check with margin and total sales. Keep a note of each source’s limitations. Tell decision-makers what each number measures and what it does not.
Transparency about limitations gives credibility.
Decision table
| Source | Tends to show | Caution |
|---|---|---|
| Ad platform | Conversions attributed to its activity | Longer windows and view-throughs |
| GA4 | Cross-channel attribution by its own rules | Depends on consent |
| Database | Real orders and revenue | No attribution by channel |
| Surveys | Declared source | Memory and bias |
Common mistakes to avoid
- Expecting equal numbers. Different methodologies give different results.
- Choosing the most favourable source. It biases decisions.
- Ignoring the database. It is the final reference for revenue.
How to measure the result
Track the ratio between reported conversions and real orders, by source.
Limits and exceptions
Attribution models and windows change, and consent reduces visibility. No model measures causality. Use attribution as an indicator, not as truth.
Illustrative example
Illustrative example: a store compares conversions reported by the platform, GA4 and the database. It uses the database as the total reference, GA4 to compare channels with the same rule and the platform for optimisation. It tracks differences and investigates when they widen.
In practice: checkpoints
Create a sheet with the conversions reported by each source, per week, along with real orders, and calculate the ratios. Always use the same model to compare channels and another rule for optimisation within each platform. When differences widen, look for the cause: consent, tags, windows or new campaigns. Record conclusions, and agree with management which source is the reference for each type of decision. Run periodic holdout tests to check whether the attributed conversions are truly incremental, and use their results to adjust how much you trust each source.
Frequently asked questions
Which source should I follow?
Use the CRM for real results and the platforms for optimisation.
Why does GA4 show fewer conversions?
Because of different models, consent, blockers and windows.
How do I validate the real effect?
With incrementality tests and period comparisons.
Which is the right source?
None is perfect; use each for what it serves and compare with the database.
How do I decide budgets?
With incremental evidence (tests) in addition to attribution.
Understand why GA4 and platforms differ, use the database as total reference and validate important decisions with incrementality tests.
Sources and recommended reading
Last editorial review: October 2026. Informational content; always confirm current conditions in the official sources before deciding.