Google has made its Meridian GeoX incrementality tool generally available worldwide and added a first-party data uplift metric. The tools are worth having, but the scorekeeper also sells the ads, so Australian marketers should keep ownership of the answer.
The Take: Google has pushed its incrementality tools out to every advertiser, which is useful and self-interested at once. Free measurement from the company selling the ads is not neutral measurement. Take the tools. Keep the scoring in your own hands.
The tool: Google has made Meridian GeoX generally available worldwide, moving its open-source geo-experiment tool out of beta so any advertiser can run causal tests on whether media actually drove sales. It also added a Data Strength Uplift metric that scores how good a brand's first-party data setup really is. The pitch is that a marketer can finally prove which media drove sales rather than argue about it over a slide deck.
The counter: Geo experiments are the honest end of measurement. You hold media back in some regions, run it in others and read the difference, rather than trusting a platform to tell you which clicks it deserves credit for. The catch is asking one company to design the experiment, supply the data and read the result. Google's own figures say advertisers who moved to its tag gateway saw 14% more conversions, which is exactly the sort of platform-reported number an independent geo test exists to check.
The conversion lift Google reports for its own tag gateway, the kind of platform claim a geo experiment is built to verify
The fix: Australian marketers lost most of their cheap third-party data to privacy changes, so measurement that leans on clean first-party data is the right direction. Ownership of the answer is the part to protect. Independence is not paranoia here, it is basic hygiene when the scorekeeper also sells the game. A tool being open source and running on your data is genuinely worth having, provided you control what it is measured against.
For Australian operators: Use Meridian, but run the geo experiments against outcomes you own, like sales and signups, not platform-defined conversions. Keep a copy of the raw data outside Google so you can re-run the numbers yourself. If a measurement tool only ever tells you to spend more with the company that built it, you have your answer about how far to trust it.