---
title: "Facebook Ads vs GA4 Conversions: Why the Numbers Differ"
canonical: "https://www.owox.com/blog/articles/differences-facebook-ads-google-analytics-4"
updated: "2026-10-29"
---

# Facebook Ads vs GA4 conversions: why the numbers differ and how to read them

Meta and GA4 will not report the same conversions. See which gap stays constant, which one swings, the documented causes, and what to compare instead.

[GA4](/blog/topics/ga4) · Updated October 19, 2026 · [Vadym Kramarenko](/team/vadym-kramarenko) · 7 min read

![Facebook Ads vs GA4 conversions: why the numbers differ and how to read them](https://cdn.owox.com/www/content-marketing-article/differences-facebook-ads-google-analytics-4/thumbnail/3-6eji31.png/public)

**Facebook Ads and Google Analytics 4 report different conversion numbers because they count different things by different rules.** Meta credits conversions to ad interactions inside its own attribution window. GA4 counts the sessions that reached your site and credits key events by its own model. A gap is expected. The useful question is which part of it holds still, because that part you can monitor, and which part swings, because that part you cannot correct.

Here is the same year of Facebook paid traffic in the \[Demo\] E-commerce store, month by month. The demo store has no GA4 property, so its sessions and orders stand in for the numbers GA4 would record. On the left, sessions recorded for every 100 Facebook clicks. On the right, how many more conversions Facebook reported than the store had completed orders.

![Two bar charts of Facebook paid traffic by month in 2025. Sessions per 100 clicks sits between 83 and 86 in every month. Facebook-reported conversions exceed completed store orders by between 35% and 112%, changing from month to month.](https://cdn.owox.com/www/content-marketing-article/differences-facebook-ads-google-analytics-4/body/chart-1-facebook-gaps-by-month.png/public)

The click gap is a constant here: 83 to 86 sessions per 100 clicks, every month. In a real account expect it to move a little, and read a sharp move as the signal. The conversion gap is not: it ran from +35% in July to +112% in August. The rest of this page explains why, and what to do about it.

## The click gap holds still, so use it as a health check

Meta’s reports count clicks in two ways. According to [Meta’s Help Center](https://www.facebook.com/business/help/787506997938504), “clicks (all)” includes [link clicks](/metrics/link-clicks) plus reactions, comments, shares, clicks to the Page profile and clicks that expand media to full screen, so it can be larger than link clicks. And a click only becomes a session if the page loads and the Google tag fires: a visitor who closes the page first is a click to Meta and nothing to GA4.

In the store data, that loss is steady at 14 to 17 clicks in 100. That makes it a monitor. If the ratio drops for a week, something broke: a tag, a redirect, a consent banner, a slow page. A ratio that is steady tells you the tracking is working, and it is the one gap here that a reconciliation can use.

| Month, 2025 | Facebook clicks | Sessions | Sessions per 100 clicks | Facebook-reported conversions | Completed store orders | Facebook reports more by |
| --- | --- | --- | --- | --- | --- | --- |
| Jan | 4,018 | 3,340 | 83 | 62 | 35 | 77% |
| Feb | 3,836 | 3,206 | 84 | 75 | 36 | 108% |
| Mar | 5,012 | 4,152 | 83 | 82 | 55 | 49% |
| Apr | 5,266 | 4,517 | 86 | 90 | 63 | 43% |
| May | 5,716 | 4,787 | 84 | 108 | 66 | 64% |
| Jun | 5,851 | 4,962 | 85 | 119 | 59 | 102% |
| Jul | 5,701 | 4,804 | 84 | 84 | 62 | 35% |
| Aug | 5,783 | 4,827 | 83 | 104 | 49 | 112% |
| Sep | 6,445 | 5,399 | 84 | 96 | 52 | 85% |
| Oct | 6,938 | 5,861 | 84 | 134 | 85 | 58% |
| Nov | 7,727 | 6,519 | 84 | 121 | 82 | 48% |
| Dec | 8,283 | 6,969 | 84 | 139 | 82 | 70% |

## The conversion gap swings, so do not correct it with a factor

Across the year Facebook reported 1,214 conversions against 726 completed store orders, 67% more. But no month sits at 67%, and a constant correction factor would be wrong in every one of them. These are the documented reasons the two sides count differently.

|  | Meta (Facebook Ads) | GA4 |
| --- | --- | --- |
| Attribution windows | Click-through 1-day or 7-day, view-through 1-day, engage-through 1-day. The default is 7-day click, 1-day view, 1-day engagement | Lookback of 30 days for acquisition events (7 can be chosen) and 90 days for other key events (30 or 60 can be chosen) |
| Credit for ads that were not clicked | Counts view-through and engagement within one day | Google describes click credit and, for YouTube, engaged views. It does not describe view-through credit for Meta ads |
| Model | Its own attribution | Data-driven by default, or paid and organic last click, or Google paid channels last click |
| Source of the traffic | The ad platform knows its own ads | Needs UTM tags on non-Google ads, or the traffic arrives without a campaign |
| Tracking | A Pixel that ad blockers can stop; Conversions API for server events | The Google tag; a session exists only if it loaded |
| Modelled results | Meta says statistical modelling may be used for some results | Reports from the tag’s own data |

Sources, read in October 2026: Meta’s Help Center pages on [attribution settings](https://www.facebook.com/business/help/460276478298895), [results](https://www.facebook.com/business/help/611432918970668) and [why numbers differ](https://www.facebook.com/business/help/147965221941551); Google’s help pages on [attribution models](https://support.google.com/analytics/answer/10596866), [how credit is attributed](https://support.google.com/analytics/answer/12958241) and [lookback windows](https://support.google.com/analytics/answer/10597962).

Some of the older explanations no longer apply as stated. [Meta’s developer documentation](https://developers.facebook.com/documentation/ads-commerce/marketing-api/insights/best-practices) says that, since 10 June 2025, off-Meta actions such as web purchases are reported at conversion time, and [GA4’s Key event attribution models report](https://support.google.com/analytics/answer/10596865) uses Event time by default. Do not assume the two disagree about the date: check the settings in your own account. And [Meta announced](https://developers.facebook.com/blog/post/2025/10/16/ads-insights-api-metric-availability-updates/) that its Ads Insights API stops returning the 7-day and 28\-day view windows from 12 January 2026, so a view-through column you used to export may no longer arrive.

Three more causes are worth checking first because you control them:

1.  **UTM tags.** [Google’s documentation](https://support.google.com/analytics/answer/11242870) says non-Google ad traffic has to be tagged manually with UTM parameters, and that missing UTMs show up as `(not set)`. The Facebook conversions that belong to untagged clicks have no campaign on the GA4 side.
2.  **Pixel and Conversions API.** Meta recommends running the [Conversions API alongside the Pixel](https://developers.facebook.com/documentation/ads-commerce/conversions-api/deduplicate-pixel-and-server-events). To avoid counting an event twice, both copies need the same `event_name` and the same `event_id`.
3.  **iOS and modelled results.** For people on iOS 14.5 or later, Meta says website conversion campaigns [can use Aggregated Event Measurement](https://www.facebook.com/business/help/331612538028890), and its results page says statistical modelling may be used for some results. Part of the Facebook number may therefore not be a count of recorded events.

## The Turning Point

OWOX Data Marts

See your first report built in real time. _15 minutes._

1.  Connect your data warehouse
2.  Pick your metrics
3.  Get a live Google Sheets report

In the time it takes to write a ticket. Then imagine never writing that ticket again.

[Book a Demo](/book-a-call)

We'll use your actual use case

## Compare like with like, and say which number decides what

You will not make the two sides match, and a page that promises you can is wrong. Do this instead:

*   **Decide budget on your own orders.** Meta’s number is the only one that exists at ad level, so use it to compare creatives and audiences inside Meta. Use completed orders, in GA4 or in your store, to decide how much each platform gets.
*   **Keep both numbers in the report, with the ratio.** A column for Facebook conversions, one for store orders and one for the gap shows the swing instead of hiding it.
*   **Watch the click ratio.** A steady sessions-per-100\-clicks figure is your tracking health check.
*   **Tag every ad.** Meta traffic gets a campaign in GA4 only through UTM parameters. [Google’s campaign data import](https://support.google.com/analytics/answer/10071305), which was called cost data import, has a direct integration for Meta, so cost can sit next to sessions in GA4 if you want it there.

This is the query behind the table. Hand it to whoever owns your warehouse: it reads the demo store’s views, so they swap in your own ad performance, sessions and orders tables.

```
WITH ad AS (
  SELECT FORMAT_DATE('%Y-%m', date) AS month, SUM(clicks) AS clicks,
         SUM(platform_conversions) AS platform_conversions
  FROM `owox-demo.ecommerce.ad_performance`
  WHERE date BETWEEN '2025-01-01' AND '2025-12-31' AND source = 'facebook' AND medium = 'cpc'
  GROUP BY month
),
ss AS (
  SELECT FORMAT_DATE('%Y-%m', date) AS month, COUNT(*) AS sessions
  FROM `owox-demo.ecommerce.sessions`
  WHERE date BETWEEN '2025-01-01' AND '2025-12-31' AND source = 'facebook' AND medium = 'cpc'
  GROUP BY month
),
od AS (
  SELECT FORMAT_DATE('%Y-%m', o.order_date) AS month, COUNT(DISTINCT o.order_id) AS store_orders
  FROM `owox-demo.ecommerce.orders` AS o
  JOIN `owox-demo.ecommerce.sessions` AS s ON s.session_id = o.session_id
  WHERE o.order_date BETWEEN '2025-01-01' AND '2025-12-31' AND o.status = 'Completed'
    AND s.source = 'facebook' AND s.medium = 'cpc'
  GROUP BY month
)
SELECT month, clicks, sessions, ROUND(sessions / clicks * 100) AS sessions_per_100_clicks,
       ROUND(platform_conversions) AS platform_conversions, store_orders,
       ROUND((platform_conversions / store_orders - 1) * 100) AS pct_more
FROM ad JOIN ss USING (month) JOIN od USING (month)
ORDER BY month
```

Store orders here are attributed by the source of the order’s own session, a simple rule. To report this on your own data every month, load the Facebook Ads data into your warehouse and join it to your orders once. How a declared join works is covered in [AI-ready marketing analytics data](/blog/articles/ai-ready-marketing-analytics-data), and [OWOX Data Marts](https://www.owox.com/app-signup) can hold that definition so every report reads the same one.

## FAQ

## Frequently Asked Questions

What are the common reasons for inconsistencies between Facebook Ads and Google Analytics 4 data?

Different attribution windows and models, a Meta Pixel that ad blockers can stop, missing UTM tags on the GA4 side, view-through and engagement credit that Meta counts and GA4's documentation does not describe, and modelled results in Meta's numbers. In the demo store, sessions stayed between 83 and 86 per 100 Facebook clicks every month, while Facebook's conversions exceeded store orders by anything from 35% to 112%.

How do I track Facebook ads in Google Analytics 4?

Tag every ad's URL with UTM parameters: utm\_source, utm\_medium and utm\_campaign. Google's documentation says non-Google ad traffic has to be tagged manually, and that missing UTMs show up as (not set). Then read the traffic in GA4's Traffic acquisition report by session source / medium and campaign.

What steps can I take to minimize discrepancies between Facebook Ads and Google Analytics 4 data?

You cannot make them match. Tag every ad, run the Conversions API alongside the Pixel with the same event\_name and event\_id, compare like with like, and keep both numbers in the report with the ratio between them. A steady sessions-to-clicks ratio shows the tracking is healthy; the conversion gap will still move from month to month.

How do I analyze Facebook ads with Google Analytics?

Tag every Facebook ad URL with UTM parameters, then read the traffic in GA4's Traffic acquisition report by session source / medium and campaign. To see spend next to it, import Facebook ad costs through GA4's campaign data import, which has a direct Meta connection. The Meta Pixel and GA4 are separate and are not connected to each other.

Why is Google Analytics showing incorrect data?

A GA4 number can be right and still differ from the ad platform's. Check that the tag fires on every page, whether UTM parameters are missing (they show as (not set)), which attribution model and lookback window the report uses, and the date range. Then compare it with orders in your store.

What is the discrepancy between Google Analytics and Facebook Ads?

Facebook counts several types of clicks and credits conversions inside its own attribution windows, including view-through and engagement. GA4 counts sessions and credits key events with its own model and lookback window. Neither is wrong: they count different things.

## Who wrote this

![Vadym Kramarenko](https://cdn.owox.com/www/webflow/6798e703afdf787334941cde_Vadym-Kramarenko.png/public)

[Vadym Kramarenko](/team/vadym-kramarenko) Growth Marketing Manager

Vadym Kramarenko is a Growth Marketing Manager at OWOX, where he drives user acquisition and product-led growth strategies. He hosts the OWOX podcast, interviewing analytics professionals about data-driven marketing, attribution, and reporting best practices. Vadym specializes in turning complex analytics concepts into practical, actionable marketing frameworks.

[LinkedIn](https://www.linkedin.com/in/vadim-kramarenko/) · [All articles](/team/vadym-kramarenko)

[GA4](/blog/topics/ga4)

## Related articles

[GA4 · GA4 Default Channel Group: The Rules, and How to Build Your Own · October 7, 2026](/blog/articles/channel-grouping-in-google-analytics4-vs-universal-analytics)

[Marketing Strategy · What is CTR, and What is a Good Click-Through Rate? · October 7, 2026](/blog/articles/what-is-ctr)

[GA4 · GA4 Measurement Protocol: Send an Event and Debug It · October 19, 2026](/blog/articles/ga4-measurement-protocol)

[See all articles →](/blog/articles)

## Links

Pages this page links to, on this site and on docs.owox.com. Where the page has a Markdown twin, its address follows the link.

- [GA4](https://www.owox.com/blog/topics/ga4) — /blog/topics/ga4.md
- [Vadym Kramarenko](https://www.owox.com/team/vadym-kramarenko) — /team/vadym-kramarenko.md
- [link clicks](https://www.owox.com/metrics/link-clicks) — /metrics/link-clicks.md
- [Book a Demo](https://www.owox.com/book-a-call) — /book-a-call.md
- [AI-ready marketing analytics data](https://www.owox.com/blog/articles/ai-ready-marketing-analytics-data) — /blog/articles/ai-ready-marketing-analytics-data.md
- [OWOX Data Marts](https://www.owox.com/app-signup)
- [GA4 · GA4 Default Channel Group: The Rules, and How to Build Your Own · October 7, 2026](https://www.owox.com/blog/articles/channel-grouping-in-google-analytics4-vs-universal-analytics) — /blog/articles/channel-grouping-in-google-analytics4-vs-universal-analytics.md
- [Marketing Strategy · What is CTR, and What is a Good Click-Through Rate? · October 7, 2026](https://www.owox.com/blog/articles/what-is-ctr) — /blog/articles/what-is-ctr.md
- [GA4 · GA4 Measurement Protocol: Send an Event and Debug It · October 19, 2026](https://www.owox.com/blog/articles/ga4-measurement-protocol) — /blog/articles/ga4-measurement-protocol.md
- [See all articles →](https://www.owox.com/blog/articles) — /blog/articles.md
