ecommerce

Ecommerce update from Google I/O

This is a great read if you work in ecommerce. Every page becomes important.

Could Gemini overtake ChatGPT by year-end? Possibly. Similarweb data shows ChatGPT’s Gen AI traffic share fell from 77.6% to 53.7% in 12 months, while Gemini rose from 7.27% to 26.7%. If that trend continues, Gemini could get very close by late 2026.

Ecommerce AI Search Optimization: What Citation Patterns Across 5 Subverticals Tell Us About Optimizing Beyond PDPs and PLPs by @aleyda aleydasolis.com/en/ai-search/e…

We put together a directory of ecommerce MCP servers, covering Shopify, Stripe, WooCommerce and more. Useful if you’re exploring how AI assistants can plug into the commerce stack. ecommerceguide.com/mcp/

Big ecommerce experience update from Google.

This is wild. If you run discount or coupon codes, this is a must-watch. Exposing Honey's Evil Business Model youtu.be/wwB3FmbcC88 Bravo @megalag 👏

Hat-trick of official Google responses on SEO and GEO. July 2025: Gary Illyes, Analyst on the Google Search team. September 2025: Danny Sullivan (former Google Search Liaison). December 2025: John Mueller, Google Search Relations.

Big AI shopping update from Google. x.com/rmstein/status…

A little helpful update from the @googlesearchc team.

@carlhendy Very good! 👏

Feels refreshing to see creativity in ecommerce that’s actually built for conversion. And not a single mention of AI. Website: In Common With

If you work in eCommerce, this should be your new favourite free tool for competitor research. This tool allows you to quickly develop your own Google Shopping Experience Scorecard and compare against top domains for your region. The tool puts to work metrics that Google use as an input to rank stores within Organic Shopping results by uncovering the relevant datasets in quite a smart and effective way. Another major benefit of the tool is that it includes its own score (out of 100) to show where a store rates overall. This is especially useful if you already have the Top Quality Store badge or are close to achieving it. As an example, JB HI-FI appear to be very close to achieving the TQS badge at a score of 82, whereas Harvey Norman still has a lot of work to do with a score of 51, being quite far off from achieving the associated benefits. For JB HI-FI, it appears as though there could be some room for improvement within the 'website quality' category. There will be clearer hints as to where the gaps are within the 'store quality' report in GMC Next, but it likely relates to comparative Core Web Vitals metrics, or something like PDP image resolution. For Harvey Norman, there look to be some foundational aspects that are holding them back related to shipping, returns, and pricing not being competitive enough. This may be harder to influence, so they would likely look to improve their Seller Rating (which is quite low), and then first look to an aspect such as 'returns', which might be a more reasonable internal discussion to have to start – considering the industry benchmark is clearly 30-day returns for most items. Oftentimes, sites will have some form of a returns policy in place, but it is not detailed (either intentionally or unintentionally) on their website. This is something that you can't skirt around, as Google validates this information itself through their Shopping Bot, with the information that is inputted within GMC Next needing to match up with what is detailed on the website. Highly recommend bookmarking this free tool for safekeeping for a time when you want to complete a competitor analysis across core metrics within your segment. Link in the comments.

Attribution is getting blurrier every day, but one of the simplest tools for understanding audience behaviour is probably gathering dust: the humble “Where did you hear about us?” pop-up. Most still list outdated search or social platforms - none mention LLMs or AI-search.

A huge problem with AI ecom is that there are too many hallucinations about what the product does/offers, that you need to verify it first-hand anyway. This problem is not solved by the new Agentic Commerce Protocol, either. And a lot of traffic is just top-of-the-funnel.

LLMs: Fewer clicks ≠ higher-quality converting traffic A $20bn ecommerce revenue study says otherwise. - 12 months. - 973 websites. - First-party data. - 164 million transactions. > papers.ssrn.com/sol3/papers.cf…

@ChatGPTapp @Shopify More details: openai.com/index/buy-it-i…

New in @ChatGPTapp: Instant Checkout, powered by the open Agentic Commerce Protocol. Now U.S. users can shop directly in chat starting with @Shopify merchants coming soon.

This issue looks to have been resolved now FYI. Just in time to be able to use this awesome free tool for comparing Top Quality Store metrics by @carlhendy: storeratings.co