Why GA4 and Google Ads never match, and how to report both

GA4 and Google Ads will never show the same conversion numbers, even when both track the same form. They answer different questions. Google Ads asks which of its own ad clicks led to a conversion and books the result against the click date. GA4 asks which channels contributed to a conversion across the whole site and books it against the day it happened.

The practical answer is to stop trying to make them match. Use each system for the question it answers well, keep them in separate tables, and investigate only when the gap between them changes suddenly. That change is almost always a tracking fault.

Click time and event time

Google Ads reports conversions against the date and time of the click that led to them. If someone clicks on 28 March and requests a demo on 4 April, the conversion appears in March. GA4 reports the same demo request on 4 April.

For B2B SaaS, with sales cycles measured in weeks, this matters a great deal. Last month's Google Ads conversions keep rising for up to the length of the conversion window, while GA4's figure for last month is fixed once the month ends. Google Ads offers "by conversion time" columns if you need an event-time view inside the platform, but the default columns are click time.

This is why we report a market's primary action (for example trial requests) in event time for anything that concerns the business, such as a brand incrementality test, and keep Google Ads' click-time numbers for optimising campaigns.

Attribution and who gets credit

Both systems offer data-driven attribution and last-click options. First click, linear, time decay and position-based models were retired in both. Data-driven attribution is the default for most Google Ads conversion actions.

The difference is the pool of channels competing for credit. The Google Ads tag only sees Google Ads interactions, so a conversion preceded by one Google Ads click gets full credit in Google Ads. GA4 can share that same conversion with an organic visit, a LinkedIn ad and an email. GA4 also excludes direct visits from credit unless the path is entirely direct. Paid search usually looks smaller in GA4 as a result.

GA4 offers three reporting models: data-driven, paid and organic last click, and Google paid channels last click. Check which one your property uses before comparing reports, because a change here moves every channel's numbers.

Conversion windows

A Google Ads click-through conversion window defaults to 30 days and can be set from 1 to 90 days per conversion action. GA4's lookback window defaults to 90 days for most key events (30 or 60 are the alternatives) and 30 days for acquisition events. Changes to conversion and lookback windows apply going forward only.

A demo request 45 days after the click counts in GA4 under default settings and is missed by a Google Ads action left on 30 days. In long-cycle B2B, we usually set Google Ads windows to match the real time from first click to conversion.

Consent and modelled conversions in Europe

In the EEA, Consent Mode v2 passes two further signals, ad_user_data and ad_personalization, alongside the storage signals. When a visitor declines, both systems can model some of the missing data, but under different rules.

• Google Ads models conversions when consent mode is implemented correctly and the account passes 700 ad clicks over 7 days per country and domain grouping. Modelled conversions appear in the Conversions column and can take several days to stabilise.
• GA4 behavioural modelling needs at least 1,000 events a day with analytics storage denied for 7 days, plus at least 1,000 daily consented users on 7 of the previous 28 days. Modelled data only shows with the Blended reporting identity and is not included in BigQuery export.

In the accounts we run, B2B SaaS sites often meet the Google Ads threshold in their largest market and miss the GA4 threshold. Google Ads then fills in declined-consent conversions and GA4 does not, so the gap is widest in countries with low consent rates.

Imported GA4 conversions and native Google Ads tags

You can bid on GA4 key events imported into Google Ads, or on the native Google Ads conversion tag. Imported GA4 conversions are still reported by click date in Google Ads, can take several hours to arrive, and pass through Google Ads' invalid click filtering, so they will not equal the GA4 count either.

The bigger risk is having several versions of the same action. In audits we often find conversions split across GA4 imports, legacy Universal Analytics imports, CRM form events and old thank-you-page tags, sometimes with different primary actions in each Google Ads account. No report can be trusted until each action has one primary source and everything else is set to secondary. For most B2B accounts the next step is importing CRM stages; see our guide to offline conversion import from HubSpot and Salesforce.

When "Unassigned" jumps: a diagnosis

GA4 uses "Unassigned" when no channel rule matches an event's traffic source. A small, steady amount is normal. A sudden jump almost always means a tracking fault.

A typical pattern: after a site change, unassigned trial events jump from a handful a day to a large share of all trial requests, while organic and direct fall by a similar amount. Paid search often holds steady, because Google Ads attribution through the click ID is unaffected. From that day, the organic and direct shares in GA4 are understated.

The diagnostic order we follow:

1. Tag Manager version history. Find the container version published on or near the date of the jump and compare it with the previous one. Most cases start here.
2. Event order and session start. Check whether the conversion event fires before the Google tag has initialised the session, for example on a form in an iframe or a new thank-you page.
3. UTM and gclid handling. Look for redirects, consent banners or single-page app routes that strip query parameters before the Google tag reads them.
4. Server-side or Measurement Protocol events. Events sent without a valid client ID and session ID cannot join a session and land in Unassigned.
5. Channel rules. Non-standard utm_medium values fall outside the default channel group.

Reconciliation table

The effect sizes below are rough, qualitative ranges from the accounts we run.

The reporting rule: one source per question

Never mix Google Ads and GA4 numbers in one table. A column of platform conversions next to a column of GA4 sessions invites a division that means nothing. Pick one source of truth for each question and state it in the table heading.

A recommended reporting layout

1. Platform view, for optimisation. One section per ad platform, using its own conversions and cost, reviewed weekly.
2. GA4 view, for channel mix. Primary action by channel in event time, reviewed monthly, with Unassigned shown as its own row so a spike is visible.
3. CRM view, for pipeline. Opportunities and closed revenue by source over a rolling 90 days, which suits B2B sales cycles.

Each section carries its own heading, date basis and definition. Commentary is written from the same data as the tables so the two cannot drift apart.

Across several European markets

Consent rates, and therefore modelling, differ by country, so the gap between GA4 and Google Ads can be small in one country and large in its neighbour. Compare each market with itself over time and keep currency and time zone settings consistent across accounts and properties.

A first step

Write down, for each report you send, which system each number comes from and whether it is click time or event time. Most confusion disappears once every table carries that label. If you want a reporting setup built this way, see our dashboards and measurement services.