Shopify checkout agents 🛒, AI Overviews citation redesign 📎, Google's DMA fight 🇪🇺

Issue 4, 30 September 2026

What changed in search and AI answers, for marketing teams, in five minutes.

Headlines & Launches

Shopify lets browser agents complete checkout (4 minute read)

Shopify extended its WebMCP tools so browser agents can read and update a shopper's checkout and place the order once the buyer confirms, with no merchant setup needed and no change to what items are in the order. If you sell on Shopify, agent checkout traffic can now convert to a sale without a human touching your site, so make sure your checkout works cleanly for bot-driven visits too.

Google rolls out AI Mode monitoring to everyone (2 minute read)

Google is rolling out AI Mode's information monitoring, previously limited to paid subscribers, to all users worldwide, letting people ask it to keep watching topics like price drops or local events and get updates over time. Search will also proactively suggest things to monitor, which means repeat visibility for brands tied to ongoing local or shopping topics, not just a one-off answer.

Deep Dives & Analysis

Desktop click rates fall as mobile clicks rise (3 minute read)

Advanced Web Ranking's Q2 2026 data shows desktop click through rates dropping at top organic positions while mobile rates climbed, reversing the pattern from Q1, though the comparison spans a period when Google was fixing a Search Console logging error. If your dashboards show a sudden CTR shift by device this year, check whether it lines up with this reporting fix before reading it as a ranking or behaviour change.

ChatGPT ad clicks vary sharply by country (4 minute read)

A Graphite report using Similarweb panel data found ChatGPT ad click through rates of 0.6% to 0.75% in the US, UK, Canada, Australia and New Zealand, but 1.21% to 1.91% in Brazil, Mexico, Japan and South Korea. Similarweb sells the tracking tool behind the panel data, and OpenAI itself says it has not published performance benchmarks, so treat these as early directional numbers if you are weighing ChatGPT ads against search or social spend.

Sponsored case study claims context fixes AI's caution on brands (11 minute read)

AnswerShare, whose founder wrote this sponsored piece, says AI models flag warnings when they find even a few unexplained negative reviews, and that adding a sourced public account of a business (customer numbers, history, responses to complaints) in an llms-full.txt file cut warning-flagged answers in one client test from 27 of 42 to zero of 40. This is a vendor's own client result, not independent research, but it points at a real mechanism worth checking: if AI answers about your brand read overly cautious, look at what context is missing for the model.

Tactics & Tools

Amazon blocked Meta's Muse because robots.txt could not (7 minute read)

Amazon blocked Meta's shopping agent Muse using its Conditions of Use, not robots.txt, because Meta's crawler documentation gives no user agent name for Muse and a court ruling in August weakened the anti hacking law Amazon had previously relied on against agent vendors. If you want to control which AI shopping agents can act on your site, check whether they publish a named user agent and know that terms of service, not robots.txt, may be your main lever.

RankZero's free AI SEO audit (Sponsor)

See whether ChatGPT, Perplexity and Google AI Overviews can find and cite your brand. Free, results in 30 seconds, no signup.

Five fixes for local visibility in AI search (8 minute read)

Google and Uberall recommend auditing each location's categories, attributes, services and booking details, keeping Business Profiles current with posts and fresh visuals, and responding to reviews to catch gaps between the profile and the real customer experience. If you run multiple locations, localise content per branch and automate routine checks, then judge results against actual bookings or calls rather than a single engagement metric.

Publishers push Congress on a bill against stealth AI crawlers (6 minute read)

Executives from Conde Nast, Hearst, USA Today and others are lobbying for the Stealth Bot Prohibition Act, which would force AI crawlers to identify themselves instead of disguising traffic to dodge blocking. If it passes, it would give publishers a clearer way to detect and block AI scrapers that currently evade robots.txt controls.

Google appeals EU order to share search history data (4 minute read)

Google has formally appealed the DMA requirement that it share users' search history with competitors by January 2027, saying it cannot do so without weakening privacy and Android security. The fight is about EU data sharing rules, not rankings, but it signals more DMA driven changes to how EU search works could still be in play.