AI shopping ramps up for Black Friday… sort of
Hi everyone - we’re deep in Q4 now and about to hit Black Friday. I hope you’re all hanging in there. To add a bit more spice to it, both OpenAI and Google have pushed forward some pretty big changes, with pretty mixed results.
OpenAI’s new shopping research tool
OpenAI has announced a shopping research feature that creates its own product roundups by reading, interpreting and ranking product data across the web. It’s essentially an AI that builds comparison guides on the fly.
It doesn’t rely on pre-written articles or curated lists. Instead, it assesses attributes, reviews and specifications to generate prompts like “the quietest cordless stick vacuum for a small apartment”.
Right now, the tool doesn’t work particularly well. It’s clearly been rushed out ahead of Black Friday. But directionally, it points very clearly to where things are heading – systems forming their own view of your products based on the information available to them.
For large-catalogue brands, this all comes back to structure. Attributes need to be complete and consistent. Reviews need to reinforce the claims you want to be known for. And the content that represents your products needs to be written in a way that models can actually understand.
There’s also an interesting implication for publishers who rely on product roundups. Even if the current version is weak, tools like this are likely to be major sources of comparison content in the future.
📝 OpenAI announcement https://openai.com/index/chatgpt-shopping-research/
Google’s agentic checkout features
Ahead of OpenAI’s release, Google has also pushed further into agentic shopping ahead of Black Friday.
Two features stand out:
- The ability to ask Google to call stores on your behalf to check stock
- The start of Google’s agentic checkout - where Google can track price changes and take action when an item goes on sale
We’re seeing adoption remain very low for now, and it’s unlikely to meaningfully affect this year’s Black Friday. But everything suggests that next year will look very different. That gives brands roughly twelve months to get their product data into proper shape.
📝 Google announcement https://blog.google/products/shopping/agentic-checkout-holiday-ai-shopping/
Adobe nearing a deal for Semrush
News broke this week that Adobe is close to acquiring Semrush for around $1.9 billion.
It’s kicked off the usual debate of “SEO is alive versus SEO is dead”, which misses the point entirely.
A few thoughts:
- Adobe buying Semrush doesn’t automatically mean anything philosophical about SEO - it may simply mean Semrush was attractively priced or filled a gap in their product suite.
- Semrush has been bloated and directionless for a long time. Adobe might fix that, or it might not. It's hard to tell from the outside.
- The more interesting angle is that the concept of “SEO” is fragmenting. SEO for Shopify is not the same as SEO for Magento. AI-driven discovery is not the same as classic SEO.
The unifying theme is discovery. It’s all discovery work shaped by the constraints of the platform you are optimising for. Whether AI makes certain routes stronger or weaker doesn’t change that.
📎 WSJ coverage https://www.wsj.com/business/deals/adobe-nears-1-9-billion-deal-for-software-provider-semrush-b239845f
AI, uniqueness and the rising value of FAQ content
Aleyda Solis shared a brilliant talk this week around algorithm shifts and AI overviews, with a central message that aligns closely with what we’ve been saying for some time - uniqueness and value are becoming the core of successful content.
Slide deck https://speakerdeck.com/aleyda/navigating-algorithm-shifts-and-ai-overviews-number-smxnext?slide=44
And here’s the LinkedIn post I wrote on the subject of using FAQs as a vehicle for unique content: https://www.linkedin.com/posts/sam-wright-17b6ab6_over-the-past-year-or-so-weve-all-talked-activity-7396472784463298560-f0WD
A few key points:
- As content becomes more portable, FAQs are becoming one of the most important assets a brand can own.
- AI systems reward information that is specific, verifiable and genuinely unique. Generic content gets ignored.
- The challenge is operational - how do you produce unique information consistently?
This is where Shopify currently has a real advantage. Knowledge Base gives you a way to manage structured FAQ content that gets fed into AI systems but doesn’t appear publicly as a standalone page. We connect this with an onsite module that surfaces the same FAQs on product and collection pages - and you can decide exactly which FAQs go where.
You end up optimising in two directions at once - giving models the unique structured information they need, while strengthening classic SEO visibility at the same time.
We’re also beginning to see this influence AI search ads. The best-performing ads match very specific queries, which means SEO and PPC are slowly merging in interesting ways.
The principle that underpins all of this is the same one we’ve held for years - Shopify should be your point of truth. It’s the central place where structured data, product content and AI-ready information come together. Think of it as a data warehouse for your products.
A notable story from Shopify
A fascinating piece this week on internal fraud investigations and dismissals inside Shopify.
It’s a reminder that behind the scenes, platforms are dealing with their own challenges.
📎 The Logic https://thelogic.co/news/exclusive/shopify-sales-fraud-investigation-firings/
Shout-out - Listabl
Given how much this edition has centred around product data and attributes, it’s worth calling out Listabl.
Listabl is doing excellent work helping brands standardise, enrich and operationalise their product data. It’s a team I’ve been hugely impressed with, and if you’re working with large catalogues, they’re very much worth a look.
📎 https://www.listabl.com/