OP-026
Today, catalogue information needs to be usable not only by human Shoppers, but also in AI-mediated product discovery and evaluation.
Accurate information on a product page may still be insufficient to establish whether the product meets what a Shopper has actually asked for.
A Shopper preparing for travel may have a clear intention but several needs to satisfy. They may want luggage that fits within cabin baggage limits, stays below a certain weight, suits the trip and fits their budget.
How much of that intent becomes explicit depends on how the Shopper searches.
A conventional search might compress the requirement into a few words:
lightweight cabin suitcase 55cm expandable
A conversational search can make more of the buying intent explicit:
“I need a lightweight hard-shell cabin suitcase that is no taller than 55 cm including the wheels, weighs less than 2.8 kg, has four spinner wheels and can expand when I need more space.”
The old distinction was simple: conventional search was often expressed through keywords, while AI conversations allowed Shoppers to describe what they wanted in natural language. That distinction is disappearing. Search engines are increasingly incorporating conversational AI search, allowing Shoppers to ask longer, more complex questions. In AI shopping, a Shopper can also express comparisons, preferences, trade-offs and several constraints in one request.
Buyer intent is not new.
What is changing is how easily a Shopper can express several parts of that intent together and ask for products that satisfy them.
That changes what the catalogue information may be required to support.
Consider the cabin suitcase.
The product is described as a lightweight hard-shell cabin case with expandable storage and four spinner wheels. Its specifications include:
Dimensions: 55 × 38 × 23 cm
Weight: 2.7 kg
Material: Polycarbonate
Wheels: Four spinner wheels
Expansion: Yes
For a Shopper looking directly at the page, there is a useful amount of information. They can read the description, inspect the specifications, look at the photographs and decide whether the suitcase deserves further consideration.
Now put the same product information into the more explicit buying situation:
“I need a lightweight hard-shell cabin suitcase that is no taller than 55 cm including the wheels, weighs less than 2.8 kg, has four spinner wheels and can expand when I need more space.”
Several parts of the Shopper’s intent now depend on that information.
The material is listed as polycarbonate, supporting the hard-shell part of the Shopper’s request.
The weight is listed as 2.7 kg.
Four spinner wheels are explicitly stated.
Expandable storage is also explicit.
But one requirement cannot be established from the information provided.
The dimensions are listed as 55 × 38 × 23 cm, but there is no indication whether the 55 cm measurement includes the wheels and handles.
Nothing is obviously wrong with the product page. The dimensions may even be completely accurate.
The problem appears because the information is now being used to support a more specific buying judgement.
AI is becoming another way Shoppers discover, compare and evaluate products. That gives the Operator another reason to check whether the product information can support the Shopper’s intent.
The question is not simply whether the product information is available.
It is whether the information can support the decisions being made with it.
What does the Operator need to establish before a catalogue is ready for AI commerce?
It would be easy to answer this by saying that AI commerce requires more product information.
Sometimes it will.
But more information is not automatically better information.
The Operator could add the colour of the suitcase lining, the dimensions of the shipping carton, the location of the factory and the number of stitches in the carry handle. Those facts might all be accurate. They still would not resolve whether the 55 cm product height includes the wheels.
The issue is therefore not the amount of information alone.
What matters is whether the catalogue contains the information needed to support the relevant buying judgement, and whether that information is clear enough to be used for that purpose.
The same issue appears with another product used on the same trip.
Consider a travel adapter.
A Shopper might want:
“A travel adapter for the UK, Europe and Australia that can charge my laptop through USB-C at 65W.”
A catalogue could contain:
Regions: UK / Europe / Australia / US
USB-C ports: 2
USB-C output: 65W
USB-A ports: 2
Maximum AC load: 2500W
On the surface, the information appears to answer the requirement.
But USB-C output: 65W still leaves an important question.
Does 65W refer to the maximum output available from one USB-C port, or to the combined output available across both ports?
The figure can be accurate while its meaning remains insufficiently clear for the buying decision being made.
Adding another paragraph of marketing copy does not settle it.
Moving the same 65W value into another technical field does not settle it either.
The underlying fact still needs to be established.
This matters because technical availability and catalogue readiness are not the same judgement.
Product information can be transferred through feeds, structured fields or other technical connections. Those mechanisms can help make information available to another system.
Technical availability does not resolve an ambiguous fact or supply a fact that has never been recorded. It also does not establish which product facts matter to the Shopper’s decision.
That still requires the Operator to examine the product, the available information and the situations in which a Shopper may need to choose between alternatives.
It also means that AI commerce does not create a reason to record every conceivable product fact.
Different buying decisions depend on different product facts.
Whether the external height includes wheels can materially affect the suitability of a suitcase.
Whether 65W applies to one USB-C port or is shared between two ports can materially affect the suitability of a travel adapter.
Those questions are relevant because they can change whether the product meets an expressed requirement.
Another possible fact may be accurate but make no material difference to the decision being considered.
The Operator therefore cannot prepare a catalogue for AI commerce simply by expanding every product record.
Nor can one fixed set of fields establish readiness for every product category.
A consistent catalogue structure remains useful. So do accurate attributes, clear specifications and dependable product records.
But consistency does not remove the need for judgement.
The Operator still has to decide which facts matter. They also need to establish those facts and record them clearly enough to support the Shopper’s intent as a whole.
That is why preparing for AI commerce is not only a technical task. It also requires Catalogue Optimisation judgement.
AI-commerce readiness requires more than technical product availability. The recorded product information must also be clear and dependable enough to support the Shopper’s request as a whole.
The Operator is reviewing the cabin suitcase before a wider AI-commerce rollout.
Instead of starting with the question:
“What extra information should we add for AI?”
they start with a realistic Shopper request:
Hard shell. Under 2.8 kg. No taller than 55 cm including wheels. Four spinner wheels. Expandable.
They open the product record and check each part of that request against the recorded product information.
They confirm the hard-shell construction from the polycarbonate specification. They check the source information and confirm that the recorded weight of 2.7 kg refers to the complete suitcase. Four spinner wheels and expansion are already explicitly recorded.
Those parts of the request are supported, so the Operator leaves that information unchanged.
Then they reach the dimensions.
The product record shows:
55 × 38 × 23 cm
That establishes the measurements but not what those measurements include.
The Operator checks the current manufacturer specification.
It states:
External dimensions including wheels and handles: 55 × 38 × 23 cm.
The original catalogue value was not numerically incorrect.
It was not clear enough to answer this part of the Shopper’s request.
The Operator updates the product information to make the scope of the measurement explicit:
External dimensions, including wheels and handles: 55 × 38 × 23 cm
They do not rewrite the entire product page.
They do not add unrelated specifications simply because an AI system may process the catalogue.
They leave the information that already supports the relevant buying decisions unchanged.
The same method can reveal a different result elsewhere.
If the manufacturer lists USB-C output: 65W but does not say whether that output applies to one port or is shared across both ports, the Operator still does not have a dependable fact. They should not turn that uncertainty into a more confident catalogue claim.
The appropriate action may be to establish the fact from a dependable source before changing anything.
That is an important part of preparing a catalogue for AI commerce.
The Operator is not trying to predict every question an AI system might receive.
They are checking whether the recorded information can support the Shopper’s request as a whole.
AI commerce introduces another setting in which product information may be used to help a Shopper choose.
It does not remove the need for professional Catalogue Optimisation judgement.
A technical connection can make a product available.
A structured field can make a fact easier to process.
Neither is enough to show that the right fact has been established or that an existing fact is clear enough.
Those remain Operator responsibilities.
This gives the Operator a more useful way to approach AI-commerce readiness.
They do not need to create a second catalogue for AI.
They do not need to add every possible product fact.
They do not need to change information that already does its job.
They need to identify what the Shopper is trying to decide. Then they can establish which product facts matter and correct the gaps or ambiguities that prevent a dependable answer.
For the Operator, AI commerce extends an existing professional responsibility. They still have to decide what product information is needed, establish it carefully and keep it clear enough to support real buying decisions. For the Shopper, the result is a stronger factual basis for determining whether a product actually matches what they asked for.