Brendon Kraham, Google's VP of Search and Commerce for Global Ads Solutions, published a piece on Think with Google this week aimed squarely at CMOs. The core message: Google does not evaluate third-party SEO tools or vendors, and those tools have no access to Google's internal metrics. His exact words: "Google does not evaluate third-party SEO tools or vendors directly, and they have no access to our internal metrics."
If you're in marketing ops, that sentence should change how you govern your SEO reporting stack.
What Google actually said (and didn't say)
Google isn't telling you to stop using Ahrefs, Semrush, or any other third-party tool. The statement is narrower and more useful than that. Google is saying that authority scores, difficulty ratings, traffic forecasts, and similar metrics from these platforms are external estimates. They aren't signals pulled from Google's ranking systems. They aren't proxies for something Google measures internally. They're the tool vendor's model, built on the tool vendor's crawl data.
Google also warned site owners to be skeptical of vendors claiming their tools are "approved" or "acceptable" by Google Search. No third-party tool carries Google's endorsement. Period.
The guidance pointed marketers toward Google Search Console as the first-party source for accurate search performance data, including impressions from AI Search features like AI Overviews. Merchant Center covers product listings. Google said these reports serve as a baseline for tracking SEO gains, with more metrics coming over time.
The ops problem: proxy metrics masquerading as KPIs
Here's where this gets practical. In a lot of B2B SaaS orgs, third-party domain authority or keyword difficulty scores quietly become the numbers leadership watches. They show up in monthly decks. They inform budget allocation. They shape hiring decisions. And none of them reflect what Google actually uses to rank content.
That's a measurement governance problem. It's not that the metrics are worthless. They can be useful for competitive benchmarking, trend monitoring, and prioritization. The problem is treating them as ground truth when they're directional at best.
Think of it like reporting platform-attributed conversions as if they were incrementality data. The dashboard says one thing; a holdout test might say something different. Same dynamic here. Your SEO tool's traffic estimate and your Search Console data will diverge. When they do, Google just told you which one to trust.
What to actually measure (and how to wire it)
The move for marketing ops is straightforward but requires some plumbing. Build your SEO measurement on first-party sources, then stitch them to pipeline.
Layer 1: Google Search Console. Queries, pages, clicks, impressions, CTR, position. This is the only data source that reflects Google's actual reporting on your search performance, including AI Search features. Treat it as the baseline.
Layer 2: GA4. Connect organic sessions to on-site behavior. Track organic signups, demo requests, and content-assisted conversions. If you're running AI Overviews traffic through the same attribution model as standard organic, flag it separately so you can spot differences in conversion rate and session quality.
Layer 3: CRM / pipeline. This is where SEO reporting stops being a vanity exercise. Map organic-sourced MQLs to SQLs to pipeline value to closed-won revenue. The whole point is answering whether organic search generates qualified pipeline, not whether your domain rating went up three points.
Third-party tools sit alongside this stack as directional inputs. Use them for keyword research, competitive gap analysis, and technical audit prioritization. Validate their recommendations against GSC data and your own conversion metrics before acting on them.
Vendor management gets a new filter
Google's statement also gives procurement teams a concrete evaluation criterion. If an SEO vendor implies that their metrics reflect Google's internal signals, or that their tool is endorsed by Google, that's a red flag. Ask vendors to disclose their methodology. Require transparency about what their scores actually model versus what they don't. Don't let a black-box metric become a KPI that drives budget.
This isn't about being anti-tool. Tools reduce manual work and surface patterns humans miss. But the ops discipline is keeping clear documentation on what each metric represents, where the data comes from, and what decisions it should (and shouldn't) inform.
The AI search wrinkle
Google's guidance also reiterated that AEO (Answer Engine Optimization) and GEO (Generative Engine Optimization) are part of standard SEO. With AI Overviews now occupying real estate at the top of SERPs, the temptation to buy a new category of AI-visibility tools is real. Some of those tools offer genuinely useful data on brand presence in AI-generated answers.
But the same rule applies. Google says these tools don't have access to its internal metrics either. Search Console is expanding its AI Search reporting, and Google says more metrics are coming. The gap between what first-party reporting covers and what teams need to measure AI search performance is real, but filling it with vendor scores you can't validate against anything creates a familiar problem wearing a new label.
Kraham's statement was aimed at CMOs, but the operational consequence lands on the people who build the reporting. The measurement stack that matters starts with what Google actually tells you, connects to what your CRM confirms, and treats everything else as what it is: an estimate.
