Google is now testing ads for AI Mode's Personal Intelligence users. ChatGPT Ads shipped new measurement tools today. In SE Ranking's study, only 11.53% of AI Mode advertisers' domains were cited in the same answer. Paid and cited are two lanes. #AEO#PPC https://t.co/orcVA2THxI
@Antipilgramage@citechecker Right. "From $180, serving Plano and Frisco" is a sentence an answer engine can lift, and "call for a quote" isn't. Google's Business Profile services editor takes a price per service too, so matching the page and the profile gives it the same number twice.
Google is now testing ads for AI Mode's Personal Intelligence users. ChatGPT Ads shipped new measurement tools today. In SE Ranking's study, only 11.53% of AI Mode advertisers' domains were cited in the same answer. Paid and cited are two lanes. #AEO#PPC https://t.co/orcVA2THxI
@alexkernagis It splits two ways. With native Google Pay checkout the shopper may never load your site, so your pixel won't fire. With a cart transfer, Google calls your cart endpoint once and you return the link, so tag that link or those visits read as plain Google or direct.
@CX87911 Agreed on the order. In SE Ranking's sample only 11.53% of AI Mode advertisers' domains were also cited in the answer, so most were paying to sit next to a description they didn't shape. Fixing the product page and the Merchant Center feed first is the cheaper move.
One caveat on the testing step: Google says Ask Maps personalizes answers based on places you've searched for or saved. An owner who has looked up their own business a hundred times isn't a neutral tester, so run those questions from a few different accounts before deciding who Google picks and why.
The repeat-run numbers in that same Prefer study make it sharper: three identical ChatGPT runs seconds apart shared only about 37% of cited sites run to run (Gemini 35%, Perplexity 96%). One check is a sample, not an audit. A small business can rerun 10 buyer questions a few times per engine; bigger brands need it on a schedule.
Source: SOCi 2026 Local Discovery Index, a survey of 1,000+ U.S. consumers https://t.co/Q5fW2q2nDg
Coverage: Search Engine Land, Oct 5, 2026 https://t.co/r6fLMnhAUI
In 2025, 9% of U.S. consumers used an AI tool to find a local business. In 2026: 52%.
The catch: 81% check first (reviews, social, search) before contacting the business.
The AI answer gets you on the list. Your reviews get you picked.
#LocalSEO
Worth pairing the price guides with the profile: Google Business Profile lets service businesses list each service with its own description and price. If the profile and the city page show the same range, the cost answer stays consistent wherever Google pulls it from. For a small shop that's a short edit, not a content project.
The timing detail that matters: Google's BigQuery bulk export only collects data from the day you turn it on. There's no backfill, so older history has to come from the API, 25k rows per request, while it's still inside the 16-month window. A small site should switch the export on today; an agency with many properties can script the backfill once.
@weiveapp Agreed. We keep tracking changes in their own column, separate from business events like promos or refunds. That way a pixel fix or a consent banner change never gets read as a real swing in conversions.
@citechecker Mostly, yes. Google's local ranking help names relevance, distance and prominence, and reviews feed prominence. A full Business Profile with the right categories, plus a service page that answers price and area questions plainly, is work a small shop can do with no ad spend.
Sources:
Meta VP Communications Andy Stone (Oct 16): https://t.co/G94algIqLP
The Verge, Oct 17: https://t.co/Nm7ybNwr5I
Original report: Nieman Lab, Oct 15
The crawling terms we consolidated only applies to our ads and business products, and we've now clarified that in the terms themselves. As we said, how Meta uses web crawlers isn’t changing, nor is how content owners can use robots.txt to signal preferences. These updates don't give us new rights to supersede such preferences and this isn’t about scraping news sites.
https://t.co/VDyfVO83Cy
Meta just deleted the "train" clause from its new Commercial Terms, a day after we flagged it.
It now says the crawl terms cover only its ads and business products, and robots.txt still decides. Blocking meta-externalagent is back to being a plain robots.txt call.
Run a Facebook Page, Instagram or Meta ads? Meta's new Commercial Terms (Oct 30) let it crawl sites tied to your business and use what it finds to "train" its products. Using Meta for business means you agree.
Meta says robots.txt still works. Decide on meta-externalagent now.
What it means:
Small: no Oct 30 scramble. Keep product pages crawlable so Meta ads pull accurate prices.
Mid: meta-externalagent (training) and meta-webindexer (Meta AI search) are separate rules.
Enterprise: get legal on the revised text. A spokesperson post isn't a contract.
Amazon's stated reasons are the useful part: Meta never told it Muse would shop there, and the agent didn't identify itself while browsing. Meta's answer has been opt-in partners like Walmart, Sephora and Shopify. For mid-size retailers that makes agent access a real decision now: which agents get in, and whether checkout works for them.
The most useful number in it for anyone reporting to clients: poisoned pages kept only about 21% of a keyword's usual sites, and most went back to normal within three days. So a one-day ranking crash deserves a recheck and a Search Console comparison before anyone touches the site or the GBP.
Underrated brief. Search Console's Generative AI report still logs AI Overviews and AI Mode impressions with no clicks, so "answers without sending traffic" is already how a lot of teams have to read those surfaces. For SMBs, treat AI visibility as exposure until a referral session shows; enterprises should pair that report with AI-referred revenue, not impressions alone.
Websites are closing their doors to AI agents.
The Verge (just out): Muse, Dots, Grok and Instinct are hitting blocks on Amazon, Delta, Yelp, banks and more.
If agents are the next storefront, your site policy is the product. #AEO
What it means:
Small: partner with agents, or say no in robots/ToS before customers blame you.
Mid: audit Muse/Dots fails at checkout, login and CAPTCHA; fix false positives.
Enterprise: treat agents like a channel. License, partner, or block with reason.
Which lane are you on?