Cross-channel conversion tracking: Measuring your campaigns' contribution

If you advertise on several channels, you face one central question: which campaign is responsible for which share of the result? Buying decisions span many touchpoints and devices, consent banners leave gaps in the data, and every ad platform credits itself with success by its own rules.

The answer lies in combining several measurement methods: marketing mix models such as Google Meridian for the cross-channel view, Enhanced Conversions and Google tag gateway for more usable signals, and new features in Google Analytics 4 that expose gaps and attribute paid clicks correctly. In this article, we show how these methods work together and how we use them to build a robust measurement setup.

Three developments make cross-channel measurement harder

1. Every platform measures differently: Google Ads, Meta and TikTok use their own tracking methods and attribute conversions by their own logic. As a result, the sum of reported conversions often exceeds the real number. Add to that visits from AI assistants such as ChatGPT or Perplexity, which many reports don't yet break out separately.
2. Privacy and consent: The GDPR and consent banners have fundamentally changed tracking. When users decline cookies, data points go missing. Attribution becomes less accurate and performance data less complete.
3. Cross-device journeys: Without a persistent user ID, a customer's journey fragments into separate sessions that appear to come from different users.

So discrepancies between tools and platforms are normal. The question is how you still make sound decisions.

Marketing mix models show what each channel contributes

Marketing mix models solve a familiar problem: every platform claims the credit for itself. They model the combined effect of all marketing activity on business results, without relying on tracking individual users.

How a marketing mix model works

A marketing mix model is a statistical model that uses historical data to calculate how much each channel drives a target metric such as revenue or inquiries. Unlike last-click attribution, it works with aggregated, channel-level data and accounts for external factors such as seasonality, competition and the economy.

Google Meridian makes marketing mix models more accessible

Meridian is Google's open-source marketing mix model. It goes beyond traditional models in three ways:

• Incorporate experiments: Results from regional tests can be used to calibrate the model, so it draws on measured impact as well as historical correlations.
• Update more often: Traditional models take months to deliver new insights. Meridian can be rerun far more frequently.
• Show how channels interact: for example, how YouTube and Search reinforce each other.

Until now, the biggest hurdle has been putting the models to use. In a Harvard Business Review Analytic Services survey cited by Google, almost 40% of marketing leaders said their companies struggle to turn marketing mix model results into business decisions. That is why Google introduced the Scenario Planner in February 2026: an interface that lets you test different budget allocations and see the expected return, no coding required.

When a marketing mix model pays off

• Many channels with an offline component: TV, print and out-of-home advertising are almost impossible to attribute through clicks.
• Long sales cycles: When weeks or months pass between first contact and closed deal, click-based models reach their limits.
• Budget allocation across channels: The model shows where additional budget contributes the most.

Keep in mind: a marketing mix model does not replace campaign-level attribution. It adds the strategic view you need for budget decisions across channels. Our page on measurement and attribution explains how we set up marketing mix models and build them into campaign management.

Enhanced conversions link late-closing deals to the first click

Enhanced conversions are among the most effective ways to get attribution right, especially with long sales cycles.

How enhanced conversions work

1. A user clicks your Google ad and lands on your website.
2. They submit an inquiry or make a purchase.
3. With the conversion, your website sends hashed first-party data to Google, such as the email address as a SHA-256 value.
4. Google matches that value against signed-in Google accounts.
5. The conversion is attributed to the original click, even if cookies were blocked or deleted.

Why long sales cycles benefit most

For complex offerings, weeks often pass between first contact and closed deal. By then, cookies are long gone and sessions have expired. Enhanced conversions base attribution on stable first-party identifiers rather than short-lived cookies. For deals that close later in sales, you feed the results back from your CRM.

We work out the legal framework with your data protection team from day one: consent, privacy policy and transmitting hashed data only. Our guide to conversion tracking in Google Ads covers how enhanced conversions are set up today and what changed in 2026.

Google tag gateway recovers signals you would otherwise lose

With Google tag gateway, Google tags load from your own domain instead of Google's servers.

Why it matters

Many browsers and ad blockers block requests to known tracking domains. When tags run through your own domain, those blocks kick in less often and your measurement becomes more complete. The data still goes to Google, just through your infrastructure.

How to set it up

You set up the gateway through your content delivery network or server infrastructure. With common providers such as Cloudflare, it takes just a few steps; alternatively, you can use a load balancer in Google Cloud. You will need access to your domain's DNS or CDN settings. The most robust setup combines tag gateway with server-side tagging. Our guide to conversion tracking summarizes what Google has published on the impact, and the article on Google tag gateway covers setup in detail.

Google Analytics 4 shows you the gaps in your data

Google has added several features to GA4 that reveal where data is missing or behaving unusually.

“(data not available)” flags data that has not been processed yet

In reports, source and medium show “(data not available)” when an advertising identifier such as the GCLID or a UTM parameter is present but the data has not been processed yet or is unavailable. The campaign then appears as “(cross-network)” and the channel as “Cross-network”. Common causes are processing time, an overrepresented user ID and, occasionally, processing errors.

What this means in practice: the value shows up more often in intraday reports than in daily ones, and event-level data can still be updated up to 12 days later. You usually cannot fix it yourself. That is why we never judge a campaign on fresh daily data while a large share is still flagged this way.

System annotations explain outliers

GA4 supports annotations in line charts. Alongside your team's own notes, Google adds system annotations whenever something on Google's side has affected the data. These cannot be edited or deleted. We add our own annotations for relaunches, tracking changes and campaign launches. Documented system changes help you interpret unusual developments in the reports more quickly.

Diagnostics show where click parameters are lost

GA4 diagnostics flag when the parameters for aggregated identifiers, GBRAID and gad_ parameters, are stripped from click URLs, typically by redirects or plugins. Once fixed, the change shows up in the property within 48 hours.

Aggregated identifiers attribute paid clicks correctly, even without consent

A user clicks your Google ad but declines the use of their data for advertising. Without further information, GA4 would attribute the visit to organic search, even though the ad drove it. The result: paid channels are undervalued, organic is overvalued and budgets end up in the wrong place.

How aggregated identifiers solve the problem

If GA4 cannot use the GCLID because the user has not granted ad_user_data consent, it falls back on aggregated identifiers. That way, GA4 still attributes paid traffic to the right source, medium and campaign, without tracking individual users.

For this to work, the GBRAID and gad_ parameters must stay intact in your click URLs. If redirects or plugins strip them, coverage drops, and GA4 diagnostics will flag it. Also link your Google Ads account to your GA4 property and keep auto-tagging switched on.

How we build a reliable measurement setup across every channel

Only by combining these methods do you get a measurement setup you can base budget decisions on.

We work on three levels

• Tactical level: Enhanced conversions and Google tag gateway capture as many signals as possible.
• Strategic level: A marketing mix model shows how to allocate budget across channels.
• Quality assurance: Diagnostics, annotations and regular checks in GA4 keep the data reliable.

No single tool tells the whole story. Only together do they give you a reliable picture.

Every stage of the buying journey needs its own measurement approach

Awareness:

• Marketing mix models to measure the impact on brand and demand
• View-through and engaged-view conversions
• Assisted conversions to capture how multiple touchpoints work together

Purchase intent and conversion:

• Enhanced conversions for accurate attribution
• Google tag gateway to capture signals as completely as possible
• Data-driven attribution in Google Ads

After the purchase:

• Importing closed deals and revenue from your CRM
• Customer lifetime value in GA4
• Cohort analysis for repeat purchases

We monitor the setup continuously

Tracking is never finished. Our recommendation: update the marketing mix model quarterly with fresh campaign data, check monthly that all key tags are firing, test different attribution windows and recalibrate the model regularly with regional experiments.

Here is a generic example of how it all fits together: an online store uses Google tag gateway to capture more checkout events, enhanced conversions with order ID and hashed email for robust attribution, Meridian to understand how TV and online advertising interact, and GA4 funnel analysis to improve the steps in between.

Ultimately, it comes down to the values you report. If you only measure revenue, you will allocate budget by revenue. Wherever possible, we pass back values that reflect contribution margin, because revenue without margin is not success.

Conclusion: combining methods is what makes your data reliable

The tools for accurate conversion tracking across all campaigns are already there. The work lies in setting them up correctly and making them work together:

• Marketing mix models such as Google Meridian deliver the cross-channel view that click-based attribution cannot. The Scenario Planner makes the results usable for decision-makers.
• Enhanced conversions link late-closing deals to the first click, even without cookies.
• Google tag gateway recovers signals that ad blockers and browsers would otherwise block.
• GA4 exposes data gaps and correctly attributes paid clicks through aggregated identifiers.

Our rule: measure first, then advertise. With a reliable measurement setup, you spend less time reconciling conflicting numbers and more time deciding where your next euro should go.