On September 10, Google published a set of measurement updates that landed without the fanfare of Google Marketing Live but carry more practical weight for day-to-day campaign management.
The headline item: a Data Strength Uplift Metric, now inside Google Ads. It estimates how many additional conversions your first-party data setup is recovering and translates that into a specific ROAS lift figure. Google cites 26% as the average across accounts connecting both offline and app data through Data Manager.
Read that slowly. Google — the company that runs the auction and earns money when your ROAS improves — is telling you that their tool delivers 26% better ROAS. You should engage with that number carefully before putting it in a budget presentation.
What the metric actually measures
The Data Strength score has existed in some form for a while, mostly as a qualitative label: Good, Fair, Poor. The Uplift Metric turns it into something more specific. It calculates the delta between conversions your current data setup detects and the estimated conversions it would recover without any first-party signals.
If your tag gateway is running, if Enhanced Conversions is active, and if you're uploading offline CRM events regularly, you'll see those contributions labeled as recovered conversions.
If your data setup is thin, you'll see how many conversions you're likely losing. That's the useful part — not the 26%, but the actual gap your specific account is carrying.
The closed loop you need to understand
Here's the structural issue: Google now provides a metric quantifying the value of your first-party data on Google's platform, measured inside Google Ads, where Google earns revenue when that data improves your performance.
This isn't unique to Google. Meta's CAPI diagnostics, Amazon's attribution, and every other platform-owned performance metric share the same conflict. The platform benefits when you believe its measurement. That doesn't make the measurement wrong, but it means it needs a second opinion.
Treat Google's 26% average figure as directional, not precise. Your actual lift depends on conversion volume, data freshness, match rate quality, and how cleanly your CRM data maps to Google's identity graph. An account uploading stale weekly exports will show a Data Strength score that flatters Google's claim more than it reflects reality.
Google also announced expanded access to Meridian GeoX in the same batch of updates. GeoX runs geographic experiments to measure incremental performance independently of platform attribution — it's the honest counterpart to the Data Strength score. Slower, harder to run, but free from the closed measurement loop. If your GeoX experiment confirms the Data Strength Uplift number, you have something worth believing. If they diverge, you have a measurement problem worth investigating.
What actually changed in the product
Two structural changes matter more than the metric itself.
First, Data Manager is now embedded inside Google Analytics and Display & Video 360, not just the Google Ads tools menu. This breaks down a real configuration problem. Previously, connecting first-party data to GA4 and DV360 meant separate setups. One data connection now flows into all three surfaces.
Second, the Data Manager API now follows the IAB Tech Lab's Event and Conversions API (ECAPI) standard. This is a bigger deal than it sounds: the same data infrastructure can theoretically connect to multiple ad platforms through a single integration, rather than a separate SDK per platform. For brands running Meta CAPI, Google Enhanced Conversions, and TikTok's Events API simultaneously, ECAPI compatibility simplifies a genuinely painful configuration layer.
In practice, this is still early. Most brands won't immediately benefit from ECAPI interoperability in a meaningful way. But it signals that Google is positioning Data Manager as universal conversion infrastructure, not a Google-only walled garden.
Four things to check this week
Open Google Ads and find Data Strength in Tools. Check your current score and look at the Uplift Metric if it's populated. If you're at Poor or Fair and you have offline conversion data sitting in a CRM, that's the gap to close first — before the 26% claim means anything to your account.
Verify that Enhanced Conversions for Web is active. This is the lowest-effort lever: it hashes customer data from form fills and sends it alongside conversion events, improving match rate without additional CRM integration.
If you're uploading offline conversions, check the upload frequency. Daily uploads improve Smart Bidding signal compared to weekly batches. The 7-day DDA cutoff confirmed in late August (conversions uploaded after 7 days are excluded from data-driven attribution calculations) makes this more time-sensitive than it used to be.
Look at Meridian GeoX if you haven't already. It's open-source, free, and the only independent measurement Google has offered that doesn't run entirely inside their attribution model. Running one geo experiment before Q4 gives you calibration data you'll want when interpreting your Data Strength score over the holiday period.
Your first-party data is almost certainly underperforming its potential. But "26% average ROAS lift" is Google's number, measured on Google's platform — and it needs a second opinion before you take it to a budget conversation.
If you want a clearer read on where your account actually stands before Q4, run a free audit at Gromerce — it surfaces data quality gaps and Smart Bidding signal issues in about three minutes.
Sources: Search Engine Land, Search Engine Journal, Google Ads blog, PPC.land, September 2026
