AP-019
Catalogue problems are often found while an agency is doing something else.
A product is being updated and someone notices a missing specification. A category is being reviewed and two products use different conventions for the same attribute. A client asks for an incorrect value to be corrected.
The immediate task seems obvious.
Correct the product and continue.
Sometimes that is exactly what the agency should do.
One incorrect value can genuinely be one incorrect value. One missing field can be the result of a single incomplete product record. There is little value in turning every contained correction into a wider catalogue investigation.
But seeing a problem on one product does not necessarily establish where that problem ends.
The same missing specification may exist throughout a range. The same inconsistent attribute may appear across products imported from one supplier. What appears to be one correction may be the first visible instance of a repeated catalogue condition.
Finding a problem tells the agency something is wrong.
It does not necessarily tell the agency how much of the catalogue is affected.
Imagine an agency receives a straightforward request.
A product weight is incorrect. The client provides the correct value, the agency checks the product and updates it.
Nothing suggests that the same mistake appears elsewhere.
The correction is contained.
The agency does not need to turn an obvious one-product error into an audit of every product weight in the catalogue.
The same is true of many ordinary catalogue problems.
A title contains a typo. An attribute has been omitted while an otherwise complete product was added.
If the problem is genuinely isolated, a contained correction remains a sensible response.
The possibility of repetition does not mean every small problem requires a wider investigation or that every small problem is part of a pattern.
Sometimes the agency really has found the whole problem.
The difficulty is that an individual product does not always provide enough evidence to make that judgement.
Consider a product missing its capacity specification.
The agency can see the field is absent and can establish what should be entered.
But the missing value itself does not show whether the omission is isolated.
It might be an isolated omission.
Or every product from the same supplier may lack the same field.
Or an older part of the category may have been created before capacity became a standard attribute.
The missing field looks the same in each case. What differs is how widely the omission occurs.
Before the agency assumes it is dealing with one correction, it may need enough evidence to know whether the occurrence is actually contained.
Suppose someone notices that a kettle records its capacity as 1500 ml while neighbouring products use 1.5L.
The problem is easy to understand.
The value can be standardised.
If the agency changes that product to match the current convention, the individual correction may be completely right.
But the correction has answered only one question:
What should this product say?
It has not established whether 1500 ml appears on one other product, another twenty or throughout an older supplier range.
From the first product, the agency can identify the type of inconsistency that exists.
It has not yet established its extent.
Consider two catalogues.
In the first, one product in a category of 200 is missing its weight.
In the second, the product where the agency notices the problem is also missing its weight, but so are most of the other products in the same range.
Viewed individually, the visible problem is identical.
One field is blank.
But they do not represent the same scope of catalogue work.
The first may require one product correction.
The second may require the agency to establish which products should contain the field, obtain or verify the information and correct a substantial part of the range.
Nothing about the original missing field became more severe.
What changed was how much of the catalogue was affected by the same problem.
If the agency defines the remediation from the first product alone, both situations can initially look like the same task.
They are not.
Correcting what the agency finds is a necessary part of catalogue maintenance. But a valid correction does not, by itself, establish whether the wider problem has been addressed.
If one inconsistent product title is corrected while another 80 products retain the same convention, the corrected product may now be right, but the same inconsistency remains across those other products.
This does not make the local correction pointless.
It means correction and measurement answer different professional questions.
A correction asks:
What should happen to this product?
Establishing extent asks:
Where else does this problem occur?
The agency may need both answers before it can say what the broader remediation actually involves.
Where repetition is plausible, the first occurrence is also evidence that similar products may need to be checked.
A missing specification in one saddle does not automatically justify inspecting every field on every product in the catalogue.
But it may justify checking whether comparable saddles use the specification consistently.
An attribute inconsistency appearing across several products from the same supplier may justify looking at the relevant supplier range.
The purpose is not to expand the scope unnecessarily.
It is to avoid treating the place where the problem became visible as proof of where the problem ends.
Establishing extent should be proportionate to the problem.
An agency does not need a complete catalogue audit every time it finds an inconsistency.
It needs enough evidence to identify the group of products in which the same issue could reasonably occur.
For some problems, that population may be a handful of comparable products.
For others, it may be a category, supplier range or group of products sharing the same attribute structure.
A catalogue-wide check may sometimes be justified, but it should follow from the nature of the problem rather than becoming the default response.
Measurement also does not require a formal score or complicated methodology.
The useful question is often much simpler:
Which products could reasonably have the same problem, and how widely does it appear among them?
That may be enough to distinguish an isolated correction from a repeated catalogue condition.
The required level of measurement depends partly on the decision the agency needs to make.
If the agency only needs to correct one urgent error immediately, the correction does not necessarily have to wait for a broader investigation.
But if the agency is about to define the scope for the client, estimate the effort or describe what completion will involve, the extent becomes more important.
A task described as correcting one missing specification is different from bringing the same specification into line across 300 products.
A request to standardise one attribute value is different from reconciling several conventions across an entire range.
The agency can only distinguish those situations once it has enough evidence of how far the problem extends.
An agency’s client may encounter a catalogue problem in much the same way the agency does: through an individual product.
The client may send a link and ask for a problem to be corrected.
That link identifies where the issue was noticed.
It does not necessarily establish the appropriate scope of the agency’s response.
If the agency checks the relevant range and finds no comparable problem, the remediation can remain contained.
If the same problem appears repeatedly, the conversation changes.
The agency is no longer discussing only the product the client happened to notice.
It can explain that the visible example is part of a wider catalogue condition and define the remediation accordingly.
The extent may affect the effort required, which products are included and what the agency and client should regard as completion.
For the agency, discovering that something is wrong and determining what that discovery means for the scope are separate decisions.
Sometimes the answer is simple: correct the product.
Sometimes the occurrence indicates that a wider group of products needs attention.
The agency cannot reliably distinguish those conditions from the seriousness of the first occurrence alone.
A problem discovered on one product can establish what is wrong without establishing how much of the catalogue needs attention.
Where repetition is plausible, establishing the relevant extent gives the agency the information needed to define the remediation rather than assuming that the first occurrence represents the whole problem.
When an agency finds a catalogue problem, it should not automatically assume either that the problem is isolated or that the whole catalogue needs investigating.
The appropriate response is proportionate.
Where the issue is clearly contained, a contained correction may be all that is required.
Where the same problem could reasonably exist across comparable products, the agency should establish how widely it occurs before treating the first occurrence as the scope of the correction.
That does not mean urgent corrections must wait while unrelated products are inspected.
It means the agency should distinguish fixing the occurrence it has found from deciding whether that occurrence represents the full scope of the catalogue problem.
The first product may be where the problem became visible.
Before defining broader remediation, the agency needs enough evidence to know whether that is also where the problem ends.