A lot of GEO, AEO, and "AI visibility" pitches are trying to sell certainty to owners who feel behind. That is exactly where bad advice gets expensive. On June 5, 2026, Google updated Search Central with direct guidance on third-party SEO tools, services, and advice. The practical reading for a building-industry operator is simple: if the vendor cannot show what official Google guidance they are following, what first-party data they checked, and what business action improves when their advice is implemented, the score is not the strategy.

That matters because many of the current AI-search promises sound technical while staying vague. They talk about AI discoverability, answer-engine positioning, citation probability, or invisible machine readability. But Google’s own framing is still grounded in normal discipline: visible helpful content, crawlability, internal links, structured data that matches visible text, and critical thinking about third-party claims.

What Google actually changed on June 5, 2026

Google’s new documentation is unusually useful because it names the exact problem. Search Central now says there is plenty of outside advice about search listings and AI experiences, sometimes called AEO or GEO, and warns that some of it misinterprets or overstates what Google says. Google also says it does not evaluate third-party SEO services and that using a tool does not guarantee ranking success.

  • Ask whether the recommendation cites official Google documentation.
  • Check whether AI-search advice aligns with Google’s generative AI guidance.
  • Use Search Console as the first-party source of truth before trusting a vendor score.
  • Treat guarantees, secret methods, and Google-approved claims as warning signs.

Why this hits building-industry businesses especially hard

A remodeler, builder, showroom, or supplier usually does not need more dashboards. They need the right next page, the right caveat, and the right workflow. If an AI answer sends a prospect to a generic overview instead of your real comparison page, or strips out the exclusion that protects a planning number from being mistaken for a final price, the business problem is operational. A citation score by itself will not catch that.

This is where the AI app-building research and current OpenAI guidance line up with Google. Repeatable AI systems are supposed to be reviewable. Good agent workflows have explicit sources, guardrails, logs, and evaluation. Your search strategy should be judged the same way. If a GEO vendor cannot explain inputs, outputs, failure modes, and the eval they use to tell a good answer from a risky one, they are selling theater.

A better test than “Did our score go up?”

Google’s June 5 hiring guidance gives a cleaner checklist than most AI-search tools do. It says to ask what results are expected, how success is measured, whether the recommendations follow Search Essentials, and whether the reasoning will be shared. For a building-industry business, that becomes a much more concrete operating review.

  • Which exact page changed, and what claim or workflow did that page support?
  • What official Google source supports the recommendation?
  • What first-party evidence did we check: Search Console, analytics, calls, discovery bookings, or qualified inquiries?
  • Did the answer preserve the caveat, comparison logic, and next action that matter to the sale?
  • Did the page lead to a real business step, not just a prettier AI visibility chart?

What not to do

Do not react to this new documentation by chasing a new fake markup layer. Google’s AI-features documentation is still explicit: you do not need new machine-readable files or special schema.org markup to appear in AI features. That means the work is still page quality, source-grounded proof, technically clean pages, and structured data that matches visible content.

It also means you should be skeptical when a vendor’s answer is "our system knows what the models want." That is not a method. At best, it is an unverified heuristic. At worst, it is a way to get you making content and schema changes that nobody can defend later.

Datum's bottom line

Use third-party GEO and AEO tools as inputs, not as authorities. The strategy is the workflow: truth-set pages, source-backed claims, fixed-query answer review, structured data that matches visible content, and business metrics tied to real next steps. If a vendor cannot show the official source, the first-party measurement, and the operating result, do not buy the score as if it were the strategy.

Sources Read

Next step, if this note maps to a problem on your desk: Discovery Call — a 1-on-1 leverage assessment for your business ($1,500 · 90 min).

Related Field Notes