Marketers struggle to defend media budgets with last-click attribution alone. Channels fragment, tracking breaks across platforms, brand touchpoints drop out of measurement. Google is expanding Data Manager and Meridian to shift focus from 'who gets credit for the conversion' to 'which investment drove business growth.'
For performance teams, affiliate managers and advertisers in iGaming, this signals a rebuild of measurement infrastructure. Not all analytics at once. First, understand where first-party data sits, how it reaches ad platforms, and what proves real incrementality rather than a polished dashboard report.
What Google updated
On September 10, 2026, Google announced new data and marketing measurement tools. Data Manager now extends beyond Google Ads: integration is added to Google Analytics and Display & Video 360. This gives advertisers more points to manage first-party data and activate it in media buying.
The second update involves the Data Manager API. Google is making the API universal and building it on the IAB Tech Lab ECAPI standard. For teams passing events from CRM, backend systems, mobile apps or partner dashboards, this is an important shift. Fewer fragmented integrations, better odds of assembling a clean signal chain.
The third change is the Data Strength Uplift Metric in Google Ads. The metric shows the incremental conversions an advertiser recovers by setting up first-party data properly. Google also adds diagnostics to Data Manager to catch data problems before they damage campaigns.
Why first-party data is central again
Advertising algorithms depend on input signal quality. If only browser events reach the system, part of the picture is lost. If offline data, app events, CRM leads and sales live separately, the marketer sees spend but not always the channel's contribution to revenue.
Google directly ties the new tools to stronger first-party data foundations. According to the company, advertisers who connect offline and app data to Data Manager see average incremental ROAS growth of 26%. For enhanced conversions, Google cites average Search conversion growth of 11% compared to standard conversion import.
These figures should not be copied into a media plan as guaranteed lift. Use them as an argument for a test. If the team still sends only basic web conversions to ad systems, the first reserve often lies not in bids and creatives but in data quality.
For the affiliate model this is especially sensitive. A CPA network, advertiser and media buyer may look at different versions of the same funnel: click, registration, deposit, repeat deposit, retention. The more gaps between these points, the easier it is to make the wrong decision about a traffic source.
Data Strength Uplift Metric: why it matters
The new Data Strength Uplift Metric is valuable because it translates the tracking conversation into business language. Usually setup of server events, enhanced conversions and CRM import competes for resources with traffic buying. The media buyer has a simple argument: another budget immediately gives volume. The analyst's answer is harder: first clean up the data.
If the metric shows additional conversion volume recovered through first-party setup, the discussion becomes concrete. You can compare: what does data improvement cost, how many conversions does it return to the system, how does ROAS change, and where does the algorithm get a cleaner signal for learning.
There is a practical sequence here:
- Separate events by value. Registration, first deposit and repeat action should not live as identical conversions.
- Check the data source. An event from site, app and CRM may have different lag and different quality.
- Verify deduplication. One sale should not be counted twice because of parallel imports.
- Watch the lag. In verticals with deferred payment or revshare, a short window often distorts ROI.
The main benefit of the metric is discipline. It forces you to link technical data setup with conversion growth, not abstract 'analytics improvement.'
Meridian: MMM with AI guidance and brand signals
Meridian is Google's open-source MMM tool. Updates add agentic capabilities: AI helps check data quality, unpack errors and guide the team through model building. Google also reports backend changes to speed up analysis runs.
For many advertisers, MMM long looked like a heavy project: econometrics, data prep, variable selection, model validation. AI guidance does not replace the analyst but lowers the entry threshold. This is useful if the team has grown to the question of budget allocation across Search, YouTube, DV360, social, offline and partner channels.
Brand signals deserve separate mention. Meridian can now include data such as Branded Google Query Volume in models. This helps assess how top-of-funnel activity affects demand later: video, TV, OOH and other formats may not close conversion immediately but lift branded search and future sales.
In performance environments, brand is often dismissed as 'expensive reach.' MMM changes the conversation. If growth in branded queries later links to sales, part of the upper-funnel budget can be defended not by CTR but by contribution to future demand. But the model requires care: bad input data will give a bad answer even if the interface became easier.
GeoX and incrementality testing
Google is also releasing Meridian GeoX globally after beta. This is an open-source library for causal geo-experiments across ad platforms. Its job is to measure incrementality by region and then feed results into MMM.
The difference between MMM and experiment matters. MMM shows relationships between marketing spend and business outcomes. A geo-experiment is closer to proving causality: in some regions activity changes, in others it stays as control, then lift is compared.
For affiliate and iGaming this logic is useful when a brand or operator buys traffic in several GEOs and debates where additional budget will produce the best result. Attribution may overvalue a channel that catches existing demand. A geo-experiment helps see whether new demand appeared because of advertising.
Practical nuance: the test must be prepared in advance. You cannot first change budgets chaotically, then call it an experiment. You need comparable regions, a clear period, stable buying rules and a metric chosen beforehand. Otherwise the result will look like analysis for the sake of a report.
What affiliate teams should do
Google's updates should be taken as a prompt toward mature measurement. You can start without major restructuring.
- Map your events. From click to deposit or purchase: what events exist, where they are stored, where they are sent.
- Check the first-party layer. Email, phone, CRM ID, app events and offline conversions should reach the system under clear consent and security rules.
- Separate CPA and revshare logic. For CPA the valid conversion matters, for revshare the quality of the user after the first action.
- Compare attribution and incrementality. A channel with high ROAS in the dashboard may not deliver the same lift in a test.
- Give brand a separate place in the model. Branded queries and deferred demand should not be mixed with bottom-of-funnel.
If the team buys traffic for offers with different payouts, it is especially important not to train algorithms on garbage signals. Registration without value, fraudulent lead and quality deposit should not feed the same goal.
Practical minimum for the next quarter
First step is a data flow audit. What events are sent to Google Ads, GA and DV360, where there are duplicates, where offline or app signals are lost. Second step is checking conversion value. If the business earns on repeat actions, the model should see at least part of downstream quality.
Next you can pick one zone for a test. For example, connect missing first-party signals and watch Data Strength Uplift Metric. Or prepare a geo-experiment through Meridian GeoX for a contested channel. You need one clean test, because it will deliver more value than ten scattered reports.
The main takeaway is simple: Google is moving advertising measurement toward a combination of data, modeling and causal verification. Teams that bring order to their events first and start counting not only conversions in the interface but real business lift will win.
