Recently I’ve been talking about the commerce catalog: how content and context are needed to meet the new demands of AI. Here, we take a step back and reflect on what I am seeing: companies are going at two speeds. The issue is one is MUCH faster than the other. Leaders vs Laggards has never been more divided.
Speed 1: Horses
The famous quote from Henry Ford: “If I had asked people what they wanted, they would have said faster horses”. We have a lot of companies that are still grooming their horses. Looking at faster horses. Considering whether their horses need some new shoes. Adding more horses.
Speed 2: New Transport
Meanwhile, other companies, let’s call them NewCo(s) are considering their businesses differently. They still have horses. Some slow, some legacy. Plus, they’re building new forms of transport. Bikes. Motorbikes, Vans, Cars. Lots of variety using a new set of building blocks like engines and wheels. LLMs and Agents. Workflow and MCP.
All companies need to be Speed 2
Of course, not all companies will be Speed 2. To have leaders, you need laggards. If you’re reading this, I encourage you to be a leader, to be speed 2. You need to develop new transport.
You don’t have to do it all at once. You certainly cannot overspend on new transport before it’s ready. Electric cars work well; however lacking a solid charging network leads to limited utility. Going all in as a commerce brand on agentic protocols, to the detriment of progress in other areas will not yield significant ROI.
So what should you do? You need to be in it, to win it. You must be learning, iterating, using. There is no watching from the sidelines. Technology is moving quickly and the advances are faster than most people realise. The latest AI models (not the free version of ChatGPT) are significantly more advanced and capable. The traffic is starting to appear from these tools in your analytics. So you must start meeting customers where they are, and learning this new world.
If I was a commerce brand, here’s what I would do to ensure I am not one of the ones watching, on speed 1:
Ensure product pages are readable by AI. This means the use of JSON-LD as a format (in the head so they don’t need to wade through the product page code) and ensure you have offers/promotions and aggregate reviews/rating included, plus global identifiers where appropriate for easier comparison
Ensure text is available: alt tags for images (long ignored by many) are now fuel for the engine, as is proper semantic use of things like H1, H2, P etc in HTML
Ensure your ecommerce platform is able to support the main agentic protocols: UCP (unified commerce protocol) and ACP (agentic commerce protocol) would be the top two. Many of the top platforms have this or have announced support soon. Know your timings: it’s higher ROI if you let the platform do it, rather than implement it yourself
Consider the content you need to produce for AI agents: OpenAI’s spec is very different to Google Shopping. Consider this in terms of content AND context, as I have written about before: an explosion of content is needed to meet the new intent-based search.
Ensure that enough content is rendered server-side to meet AI needs: if everything is loaded via Javascript, client side, it will likely be invisible to AI.
Start using AI inside your team: find some small use cases where AI can be used to make your workflow faster or higher quality. Solve cases that before were too much (human) effort, which AI can now do overnight: build your team’s AI intelligence and muscle
Many of the six points above will already be a ‘check’ for brands on the right path. Don’t ignore number six, that is where your learnings will grow fast and your brain will start to understand the world ahead.