--- type: "article" title: "#BlinkTank: Google says skip llms.txt, Shopify sends it live on every store" newsletter: "#BlinkTank" newsletter_handle: "blink-seo" newsletter_url: "https://usecommune.com/n/blink-seo" author: "@sam" published: "2026-05-20T08:15:00.000Z" canonical_url: "https://usecommune.com/n/blink-seo/a/blinktank-google-says-skip-llms-txt-shopify-sends-it-live-on" markdown_url: "https://usecommune.com/n/blink-seo/a/blinktank-google-says-skip-llms-txt-shopify-sends-it-live-on.md" chat_url: "https://usecommune.com/n/blink-seo/a/blinktank-google-says-skip-llms-txt-shopify-sends-it-live-on/chat" source_url: "https://blink-seo.kit.com/posts/blinktank-google-says-skip-llms-txt-shopify-sends-it-live-on-every-store" body_source: "imported" likes: 0 replies: 0 body_words: 1881 --- # #BlinkTank: Google says skip llms.txt, Shopify sends it live on every store ![](https://embed.filekitcdn.com/e/sH3ZoKeVgpieQqFCAeostP/gDG1fxUAgeWHdWwCuipH3) # **#BlinkTank: Google says skip llms.txt, Shopify sends it live on every store** Last issue covered Shopify’s new llms.txt file and what it’s telling AI agents about your store. This week Google has published its own guidance on AI search, and the headline is that AI search doesn’t need llms.txt, content chunking, or anything rewritten specifically for AI. On the surface that may look like a direct contradiction with what Shopify is doing, but it’s really two different things going on, and the gap between them is worth pulling apart. Also in this issue: GA4 has added a dedicated AI Assistant channel in Default Channel Groups, much to the dismay of the AI tracking SaaS market, Shopify has tightened rate limits on bots and agents, a refresh of our taxonomy article, and a piece from Ecomswap on what your DTC brand is actually worth. Finally, we are at the [National Digital Awards](https://nationaldigitalawards.co.uk/) on Friday this week as a finalist for Best SEO Agency. If you are going to be there, come and find us. ![](https://embed.filekitcdn.com/e/sH3ZoKeVgpieQqFCAeostP/mkLQRM6FXaxaB5fZWUGjQf/email) *​* On to the news! ## Google publishes its AI search guidance Google has published a new documentation page, [Optimizing your website for generative AI features on Google Search](https://developers.google.com/search/blog/2026/05/a-new-resource-for-optimizing), expanding its earlier AI features guidance. The substance is useful, particularly the parts that push back on some of the more aggressive claims circulating in the AI SEO market. ![](https://embed.filekitcdn.com/e/sH3ZoKeVgpieQqFCAeostP/wKSiUQUFgMqkfyi8tgUKrV/email) *​* The guide is explicit on three things: llms.txt and similar machine-readable files don’t get any special treatment from AI search, content doesn’t need to be chunked because Google’s systems can already read multiple topics on one page, and content doesn’t need rewriting for AI - whether to fit an AI-friendly register or to capture every long-tail variation - because the same crawl path that drives classic search results is what feeds AI Overviews and AI Mode. Google’s framing is that AEO and GEO are “still SEO”. Unsurprisingly, we broadly agree. However, a couple of editions ago we covered the llms.txt and /sitemap_agentic_discovery.xml files Shopify is now serving on every store by default. Those files are still live, and they’re doing the kind of “tell AI agents what’s important about this store” work that Google has just said AI search doesn’t need. There is obviously a tension here. However, it’s worth noting that while there’s some overlap at the moment, the.txt file has a different purpose for Shopify. Google’s documentation is about AI search inside Google itself - AI Overviews and AI Mode. Shopify’s infrastructure looks aimed at agentic commerce, where an AI agent buys something on a user’s behalf rather than answering a search query. Those are different things (even if a good chunk of the infrastructure behind them is the same), and Google’s position on one doesn’t necessarily say anything about Shopify’s plans for the other. Meanwhile, there’s a lot of shouting on LinkedIn at the moment about how important llms.txt files are, but it’s also worth saying that there’s no real evidence they’re being used for anything yet. That will change over time, but for now there’s nothing confirmed. ## GA4 adds an AI Assistant channel On 13 May, GA4 added a dedicated AI Assistant channel to its Default Channel Group. Traffic from recognised AI assistants like ChatGPT, Gemini and Claude now gets its own medium (“ai-assistant”), its own channel (“AI Assistant”), and its own campaign attribution (“(ai-assistant)”). It’s automatic with no setup required, and shows up alongside organic search, paid search, direct, and the rest. Its worth flagging though that we’re not seeing it in every account yet, so if it isn’t showing up in yours, it’s probably a phased rollout rather than anything you’ve configured wrong. This is useful as a baseline. Until 13 May, AI-driven traffic was being lumped into Direct, Referral, or Other depending on how each AI tool was sending it, which made it almost impossible to read the actual contribution of AI as a discovery channel. Having it in its own column should make it considerably easier to see whether AI is moving from “thing being talked about” to “thing actually delivering sessions” for a specific store. The caveat is that the new channel only captures the user-click side of AI traffic: someone reads an AI answer, clicks through to the site, and shows up in GA4 as a referral. The retrieval side, where an AI assistant fetches your content to generate its answer without the user ever clicking through, is still invisible to GA4 because no human visit ever happens. This was covered in a [recent LinkedIn post](https://www.linkedin.com/posts/martin-jeffrey-630577152_geoagencyleeds-aisearchoptimisation-agenticoptimisation-share-7460706416589271042-ckPu?utm_source=share&utm_medium=member_desktop&rcm=ACoAAAErnS8BD7KoDVnSAkXneNkEdiuvvYcQM18) from Martin Jeffrey - he ran server log analysis showing AI crawler retrieval running at 1.9x to 4.4x the volume of traditional organic search across the sites he looked at. This roughly matches what we’re seeing across our client base. We can’t verify it directly because Shopify stores don’t expose server logs, so this is all anecdotal, but it definitely passes the sniff test. 🔗 [Google Analytics release notes](https://support.google.com/analytics/answer/9756891)​ 🔗 [Martin Jeffrey on LinkedIn](https://www.linkedin.com/posts/martin-jeffrey-630577152_geoagencyleeds-aisearchoptimisation-agenticoptimisation-share-7460706416589271042-ckPu?utm_source=share&utm_medium=member_desktop&rcm=ACoAAAErnS8BD7KoDVnSAkXneNkEdiuvvYcQM18)​ ## Shopify tightens bot and agent rate limits We are all extremely sensitive to bot sessions in Shopify reporting at the moment. It’s worth knowing that the Shopify developer changelog on 7 May confirmed stricter rate limits are now being applied to bots and agents accessing the Storefront API and Shopify-hosted online store pages, with the bots that don’t sign their requests using Web Bot Auth subject to the strictest limits. ![Screenshot 2026-05-19 at 15.26.40.png](https://embed.filekitcdn.com/e/sH3ZoKeVgpieQqFCAeostP/sWLLHuCeBcFy5ByUCZQKux/email) *​* Web Bot Auth is a cryptographic signature merchants can generate from the Shopify admin (Online Store > Preferences > Crawler access) and hand to authorised crawling tools, which then include it on every request. Signed bots get higher rate limits, unsigned bots get throttled hardest, and signatures last up to three months before they need regenerating. The point worth keeping an eye on is that this change could explain any odd patterns showing up in your reporting from 7 May onwards. If bot session counts have shifted noticeably, or if your own tools have started returning incomplete data or unexpected errors, the rate limit change is a reasonable place to start looking. 🔗 [Shopify developer changelog: Web Bot Auth](https://shopify.dev/changelog/bots-and-agents-should-identify-themselves-via-web-bot-auth)​ ## An update to our Shopify taxonomy guide We’ve updated our guide on why taxonomy matters for Shopify stores, originally published in 2024. The argument hasn’t changed, but the evidence behind it has, and the article now reflects what we’re seeing across our client base today rather than the picture eighteen months ago. 🔗 [Why taxonomy is so important for Shopify stores](https://blinkseo.co.uk/blogs/news/why-taxonomy-is-so-important-for-shopify-stores)​ ## What your brand is actually worth Most of our conversations with founders are about getting bigger. Growth, scaling, channel mix, where the next phase of revenue comes from. What the brand might actually be worth in a sale comes up much less often, and almost always too late to do much about it. We’ve been speaking with Ecomswap recently, a sell-side M&A advisory for profitable DTC brands. Their view is that brand value comes down to EBITDA × multiple, but the multiple is anything but fixed: floor deals sit around 2x, median around 5x, and best-in-class up to 10x. The best exits start 12 to 24 months before a founder decides to sell, not 30 days after, and by the time most founders have visibility into where their multiple actually sits, the window to move it has mostly closed. They’ve put together a free valuation calculator that gives a realistic range, shows which factors are moving the multiple in which direction, and adds some market timing context. If you’ve ever wondered what the brand might sell for, it takes about five minutes. 🔗 [Ecomswap valuation calculator](https://ecomswap.com/valuation-calculator?utm_source=blinkseo&utm_medium=newsletter&utm_campaign=ecommerce_valuation_calculator&utm_content=cta_button)​ ## Jobs **Hiring:** ​[Blakely Clothing](https://www.blakelyclothing.com/) are hiring an [E-commerce Executive](https://uk.indeed.com/job/e-commerce-executive-4aaba539cea6fc54) in Norwich. Hands-on Shopify role at a fast-growing fashion brand, working alongside their eCommerce and Tech team. They’re geographically close to us and exactly the kind of large-catalogue Shopify operator we work with. ​[Elizabeth Scarlett](https://elizabeth-scarlett.com/) are hiring a [Customer Experience Lead](https://www.notion.so/3647f393e138805eb786db8527be7788?pvs=21). The team at [Blink](https://blinkseo.co.uk/) have started working Elizabeth Scarlett recently so I have a slight bias, but the framing on this one is very interesting. This is positioned as ownership of the full purchase experience end to end (pre-purchase, delivery, unboxing, transactional comms, returns, customer service), sitting in Growth as a retention lever rather than in Ops as a cost. The reason this matters now, more than it did even a year ago, is that AI search is increasingly pulling from reviews, post-purchase content, returns experience and community signals when deciding what to surface. Customer experience is becoming a direct input into marketing performance, not just an operational function. ​[Outcast](https://outcastclothing.com/) are hiring a [Head of Ecommerce](https://www.linkedin.com/jobs/view/4401649332/) to lead global ecommerce for one of Australia’s fastest-growing fashion brands. The role is on-site in Burleigh Heads, QLD. ​[Creed](https://www.creedboutique.com/) are hiring a [DTC Ecommerce Manager](https://www.linkedin.com/jobs/view/4413046297/) in London. The heritage perfumery house, founded in 1760, is going through significant change on the ecommerce side. On-site. ​[SQUATWOLF](https://squatwolf.com/) are hiring a [Head of Merchandise Planning](https://www.linkedin.com/posts/chrisbishop_were-hiring-a-head-of-merchandise-planning-share-7459972854164578305-mdAY) in Dubai, leading planning, allocation and forecasting across Ecom, Retail, Wholesale and international markets as the brand scales across the GCC and beyond. **Looking for work:** ​[Danny Wright](https://www.linkedin.com/posts/danny-wright-187a1310a_after-a-great-year-leading-e-commerce-operation-share-7460288384389996544-oGcc) is looking for his next role after a year leading ecommerce at Wowow Toys. His specialism is end-to-end growth across acquisition channels, tech, and backend operations, and he’s keeping an open mind on role type. If you are hiring and the role fits someone with large-catalogue Shopify experience, or you are looking and have that background, send me a message. ## Partner shout out: Loxa Loxa is product protection built for Shopify checkout. Customers can add cover for accidental damage, theft, and similar at the checkout, on a wide range of purchases. The Shopify-native setup means it doesn’t sit awkwardly on top of the store, and the brand gets a margin on every protected order. We’ve seen real success with clients using it across furniture, technology, and other categories where customers want some post-purchase reassurance. 🔗 [Loxa](https://loxacover.com/)​ *** If you found this useful, the best thing you can do is forward it to someone else managing a Shopify catalogue. As ever, thanks for reading. If you found this useful, I would love it if you could share it with someone you think would find it interesting - whether that is someone managing a large-catalogue Shopify store or anyone else who follows this kind of thing. See you next week. Sam *** **Who am I and what does Blink do?** I’m Sam Wright, Managing Director and co-founder of Blink SEO. We’re a search marketing agency that works exclusively with large-catalogue Shopify stores - the kind where the challenge isn’t just visibility, but helping customers find the right product in the first place. We specialise in SEO and PPC for merchants managing hundreds or thousands of products. Most of our work starts with taxonomy: how a store is structured and categorised, and what that means for organic performance, paid efficiency, and on-site discovery. If you found this useful and aren’t already subscribed, you can sign up at [blinkseo.co.uk/blinktank](http://blinkseo.co.uk/blinktank). If you run a large-catalogue Shopify store and want to talk about organic or paid performance, reply to this email or find me at [blinkseo.co.uk](http://blinkseo.co.uk). [![linkedin](https://functions-js.kit.com/icons?icon=linkedin&foreground=ffffff&background=0077b5&shape=circle&v=3)​](https://www.linkedin.com/company/blink-seo)[![youtube](https://functions-js.kit.com/icons?icon=youtube&foreground=ffffff&background=cc0000&shape=circle&v=3)​](https://www.youtube.com/channel/UCmwK1KZI2PLULLOXmBQW5fA)[![twitter](https://functions-js.kit.com/icons?icon=twitter&foreground=ffffff&background=0c85d0&shape=circle&v=3)​](https://x.com/Blink_eCommerce) Kiln House, Pottergate, Norwich, Norfolk NR2 1DX ​[Unsubscribe](https://preview.kit-mail3.com/unsubscribe) · [Preferences](https://preview.kit-mail3.com/preferences)​ *** ## Discussion No replies yet.