Aizii Learn · AI Search

[← Back to Learn](https://aizii.ai/learn)

# How to stop AI recommending your discontinued products

You stop AI from recommending a discontinued product by removing it from every source the models read: your pages, your feeds, your schema, your sitemap, your llms.txt file, and your resellers. There is no single switch. A model recommends the product because some source still says it is available, and the model has no way to know you stopped selling it until every source agrees. Fix one source and leave the rest, and the recommendation keeps coming from the source you missed.

Set expectations first. Nothing here removes a product instantly. A search-based answer changes after the next crawl of the corrected pages, and a training-based answer waits for a newer model. What you can do is make the removal complete and verifiable, so the next crawl finds nothing left to recommend.

## Why the discontinued product keeps appearing

1.  The page is still live and still says in stock. The product page was never updated, or the page was removed from the catalog but the URL still serves a page that says the product is available.
2.  The schema still says InStock. The visible page says the product is gone, but the JSON-LD still carries an InStock offer, so the model reads the markup as a declared fact.
3.  A feed still lists it. Your product feed, your Google Merchant Center data, or a direct feed to an AI assistant still carries the product as available, and the feed is fresher than your pages.
4.  Your sitemap and llms.txt file still point to it. A sitemap that still lists the old URL tells crawlers the page is current. An llms.txt file that still links to it tells models the product is part of your catalog.
5.  A reseller still sells it or still lists it. The model cites the reseller's page, which still shows the product as available, even after your own page is fixed.
6.  Old copies sit in indexes and in training. A page that was indexed months ago, or a fact learned in training, does not disappear the day you edit your site. The model keeps reading the old copy until it is replaced.

## Step 1: Decide redirect or remove for each product

Sort every discontinued product into two groups. If a close replacement exists, a newer model that answers the same need, redirect the old URL to that product page with a permanent server-side redirect. Google's redirects documentation says to use a permanent server-side redirect whenever possible, and that a permanent redirect shows the new target in search results. Redirect to the specific replacement page, not to the homepage, since the homepage does not answer the same need. If no replacement exists, let the URL return a 404 or a 410. Google's status code documentation says it does not index URLs that return a 4xx status code, and that URLs already indexed that return a 4xx are removed from the index. All 4xx codes except 429 are treated the same, so pick whichever your platform supports cleanly. Do not serve a 200 page that says "not found": a success code tells crawlers the page is live.

## Step 2: Mark the product Discontinued in schema

If you keep the page live, for example to show the replacement and preserve links pointing to it, change the availability in the Offer to the Discontinued value from schema.org. Google's Merchant Center documentation maps the Discontinued schema.org value to an out of stock feed value, so it is the value the machines already understand. Keep the rest of the markup accurate: the name, the brand, and the URL of the page the offer belongs to. Do not leave a price that suggests the product can still be bought. If you redirect the page instead, remove the Product markup from the old URL, since markup on a page that no longer sells the product creates a fact with no page behind it.

## Step 3: Pull it from every feed

Remove the product from each feed that carries it: your Google Merchant Center feed, any direct product feed you send to an AI assistant, and any feed a platform sends on your behalf. A feed that still lists a discontinued product as available will keep it in shopping results after your pages are fixed, because the feed is the fresher source. Check the feed after you edit it, and confirm the product is gone from the next export, not only from the edit screen. If your platform sends the catalog automatically, check that the discontinued product is excluded from the catalog it sends, since an automatic feed can re-add a product you removed by hand.

## Step 4: Update the sitemap, the llms.txt file, and the internal links

Remove the old URL from your sitemap and add the replacement URL if you redirected. A sitemap that still lists a discontinued product tells crawlers to keep checking a page you know is gone. Remove the product from your llms.txt file and from llms-full.txt if you keep one, and remove it from category pages and "you may also like" blocks. Then fix the internal links: old blog posts and navigation menus keep pointing at discontinued products long after the cleanup, and every live link is a signal the product still matters. For the llms.txt format, see our [guide to writing a good llms.txt file](https://aizii.ai/learn/ai-search/how-to-write-a-good-llms-txt-file).

## Step 5: Ask resellers and directories to remove it

Check the resellers, marketplaces, and directories that still list the product, and ask each one to remove it or mark it discontinued. Give them the link to your replacement product page so they can point shoppers to it. You cannot edit their pages yourself, but a reseller listing is often the page the model cites, so a fix on your site alone does not stop the recommendation. Prioritize the listings the model actually cited when you tested, since those are the ones doing the damage.

## Step 6: Verify the removal and retest

Request each old URL yourself and confirm what it returns: a 301 to the replacement, or a 404 or 410. Check that the Product schema on any page you kept shows Discontinued, not InStock. Then retest the same buying prompts that used to surface the product, and log the answer and the citations each time. If the product still appears, open the cited page: it is usually a reseller listing you missed, a feed that still carries it, or a cached copy the model has not re-read yet. There is no guaranteed removal time. Google publishes no timeframe for how fast an indexed URL drops out after it starts returning a 4xx, and no vendor promises how fast a training-based answer changes, so keep the log and retest on a schedule instead of expecting a single fix to take effect the same day.

### Say you run an outdoor gear brand that retired a tent line

Say you run an outdoor gear brand that retired a two-person tent and replaced it with a lighter model. A customer asks a model for your lightest tent, and it names the retired one, citing your old product page and a retailer's listing. You redirect the old product URL to the new tent with a permanent server-side redirect, and you remove the retired tent from your product feed, your sitemap, and your llms.txt file. You ask the retailer to update its listing to the new model. You check that the old URL returns a 301 to the new tent, and you retest the same prompt and log each result. The search-based answer changes after the next crawl. Any training-based version waits for a newer model.

### The fundamentals

These mechanics are shared across every engine and covered in full in their own guides:

-   •llms.txt: the file format, a complete sample, and per-platform publish steps, see our [guide to writing a good llms.txt file](https://aizii.ai/learn/ai-search/how-to-write-a-good-llms-txt-file)
-   •Product schema: the JSON-LD markup models read to verify price and availability in structured form, see our [product schema guide for AI search](https://aizii.ai/learn/ai-search/how-to-add-product-schema-markup-for-ai-search)
-   •Every AI crawler in one reference table, see our [robots.txt guide for AI search](https://aizii.ai/learn/ai-search/how-to-write-robots-txt-for-ai-search)

## What this guide does not cover

This guide covers stopping recommendations of products you discontinued. It does not cover correcting a wrong price or a wrong detail about a product you still sell, which is a separate guide, or the monitoring work of catching the next discontinued product before a shopper is sent to a dead page. It also does not cover the transaction layer: a recommendation that links to your store is not a completed purchase, and letting an agent finish the checkout is a separate job.

## Check your own brand now

Run the free scan. It takes seconds, no signup, and shows whether AI crawlers can read your current catalog today: aizii.ai/aeo-scan

[Run the free scan](https://aizii.ai/aeo-audit)

## Sources

-   Google Merchant Center Help, Supported structured data attributes and values (specific): [https://support.google.com/merchants/answer/6386198](https://support.google.com/merchants/answer/6386198)
-   Schema.org, ItemAvailability type (specific): [https://schema.org/ItemAvailability](https://schema.org/ItemAvailability)
-   Google Search Central, Redirects and Google Search (specific): [https://developers.google.com/search/docs/crawling-indexing/301-redirects](https://developers.google.com/search/docs/crawling-indexing/301-redirects)
-   Google Search Central, How HTTP status codes affect Google's crawlers (specific): [https://developers.google.com/crawling/docs/troubleshooting/http-status-codes](https://developers.google.com/crawling/docs/troubleshooting/http-status-codes)
-   Google Search Central, Product structured data (general docs): [https://developers.google.com/search/docs/appearance/structured-data/product](https://developers.google.com/search/docs/appearance/structured-data/product)

Note: links marked specific point to the page that supports the claim in this guide. Links marked general docs point to the vendor's general docs section for that topic, not to a page that proves every step. Check the vendor docs for the current path before relying on a menu name or click path in this guide.
