A category can remain full of product photographs after its useful buying choices have disappeared. For a merchant paying for ChatGPT ads, that creates a destination decision. Does the remaining range still justify inviting people to explore this category?
Start with the offer in the actual ad. The answer depends on whether customers can fulfil that offer with the products currently available to them. A percentage of items in stock can inform the discussion, but it cannot make the decision on its own.
Use the advertised choice as your unit of review
Consider a hypothetical store advertising an assortment of beginner sewing machines. The category contains basic machines, more advanced models and accessories. After a stock shortage, several expensive machines and many accessories remain available, while all the entry models are unavailable.
The category may still contain numerous buyable cards. That does not establish a useful beginner selection at the advertised price level. Counting replacement needles alongside machines makes the inventory picture look healthier while ignoring the buying task that prompted the paid click.
Write a short definition of the choice promised: intended use, essential features, relevant price range and market. Use facts in the ad and destination, together with choices the visitor makes on the site. This review does not require knowing the person’s private ChatGPT conversation.
Next, identify the models and variants that currently meet that definition and can be ordered in the destination’s market. Count meaningful options separately from duplicate colours or accessories. A model with only an unsuitable variant left should not count as coverage of a requirement it no longer meets.
Separate four category conditions
The following conditions are a merchandising review model, not platform statuses. They make the decision more specific than a general alert that stock is low.
The promise still holds. Enough relevant choices remain to offer the comparison described. Update claims about breadth if necessary and review how the buyable range appears on arrival. A smaller assortment can still be useful when it preserves the choices that matter.
Only a narrower promise holds. The remaining range serves one part of the original offer. A category restricted to advanced machines could support a different, accurate campaign. It should not quietly replace the beginner proposition after the visitor has clicked.
Availability is uncertain. The data cannot establish what customers can order. Assign an owner to resolve the discrepancy and decide how to handle traffic while checking. Unknown availability is not evidence that everything has sold out, nor permission to describe everything as available.
No relevant choice remains. The category may still be useful as a reference, but the current purchase invitation has lost its basis. The campaign owner should decide how to stop sending paid traffic that depends on that invitation.
Does the buyable range support the ad promise?
- The promise holds
Keep the destination and make the relevant buyable range clear.
Checkpoint - Only part remains
Narrow the ad and category together around the remaining choice.
- The data is uncertain
Resolve availability and explicitly decide how to handle traffic meanwhile.
- No relevant choice is buyable
Stop the affected purchase invitation and define restart conditions.
Checkpoint
Review the first impression of the remaining range
Open the category as a new visitor. Read its heading, introduction, main image and first product group together. An image dominated by unavailable hero products can preserve an impression of choice even when the smaller stock labels are accurate.
Make the relevant buyable range easy to find. If unavailable items remain for comparison or reference, distinguish their purpose clearly and ensure the available selection is understandable without repeated dead ends. Decide deliberately whether the page still needs those reference items in its main browsing area.
Explain material temporary shortages in language specific to the choice. A verified statement about unavailable entry models is more informative than “limited stock” when the remaining products are all in another tier. Avoid turning stock problems into default urgency copy or publishing an unconfirmed replenishment date.
Use the product stock-status guide for messages and purchase actions on individual cards or pages. The category decision here comes earlier: whether the range as a whole merits the promise being advertised. It is also separate from helping someone recover after their own filter choices return no matches.
Coordinate the offer before changing the destination
If a narrower range is worthwhile, write the revised invitation and category explanation together. In the sewing example, the store might create an accurate proposition for experienced buyers if the remaining products support it. That is a new scope for the offer, requiring deliberate campaign review.
Do not redirect the existing beginner invitation to an unrelated department simply because that department has stock. Where other products genuinely preserve the buying need, use the product alternatives guide to explain their differences. A substitution should help customers make a new choice knowingly.
OpenAI documents pause controls for ads, ad groups and campaigns in Campaign Management. Identify the affected advertising in the workflow available to your team. A category edit or an internal decision to pause is not confirmation that the relevant advertising state has changed.
Check the catalogue handoff separately
Category presentation and advertising data need their own checks. This applies especially when the merchant uses both ads that link to a category and advertising built from product feeds. Assign owners for the shop’s range, the feed records and the campaign decision, with a clear handoff between them.
OpenAI’s Product Feeds documentation calls for current availability and explains that stock updates apply asynchronously. An unavailable product stops qualifying for delivery after the change propagates. Submission alone therefore does not prove that every intended change is already reflected in advertising.
Inspect the affected records and the range the advertising actually uses after processing. Keep the source catalogue consistent with the intended stock position so later work does not restore outdated data. The product-feed preparation guide provides wider catalogue context; a successful category review still needs a separate look at the destination customers encounter.
Define recovery in terms of usable choice
Before narrowing or stopping the campaign, record what would justify restoring it. For the fictional store, that could mean a verified selection of entry machines available at the price level in the offer. “The supplier dispatched a shipment” is a different event and does not demonstrate that customers can complete an order.
Recheck the category, representative variants and the next purchase step when the recovery condition appears satisfied. Review the restored ad wording alongside the page. Record the check time, evidence and owner so that the next shortage prompts the same clear decision about the paid invitation.
Sources and scope
Assess whether the buyable category range still fulfils the paid ad promise and coordinate a narrower offer, continued traffic or a temporary stop.
Working methods and examples are editorial suggestions. Check current platform requirements and available features before implementation.