The correct handling depends entirely on whether the product is coming back. For temporarily out-of-stock items, the page should stay live, return a normal 200 status, remain indexed, and clearly communicate availability status to the user rather than being removed or redirected. For permanently discontinued products, the right move is a 301 redirect to a genuinely relevant replacement or category page if one exists, or a properly handled 404/410 if it doesn’t. Treating both situations the same way, either by taking every out-of-stock page down or by redirecting every discontinued product to an unrelated page just to “preserve authority,” causes the exact problems both approaches are meant to avoid.
Why temporary and permanent unavailability require different treatment
Google has published specific guidance on this exact scenario because it’s such a common e-commerce mistake. The underlying logic is straightforward: a page’s ranking signals (backlinks, historical click data, indexed content, accumulated relevance for its target queries) took time to build, and those signals don’t automatically transfer if you remove or redirect the page. For a product you expect to restock, removing the page or serving a 404 destroys those signals for no reason, since the product will be sellable again and you’ll want to have kept the ranking position. Worse, if a site does this repeatedly across many SKUs (taking pages down every time something goes out of stock and re-publishing when it’s back), Google’s crawlers may start treating the domain’s product pages as unstable or unreliable, which can slow re-indexing when the item does return.
For genuinely discontinued products, the calculus reverses. The page no longer represents something a user can buy, so leaving it live and indexed as-is creates a mismatch between what’s indexed and what’s actually available, which is a poor user experience and can register as a soft-404 pattern if traffic lands on a page users immediately bounce from. But the common mistake in the other direction is treating “preserve the page’s authority” as the only goal and redirecting it to something unrelated, like the homepage or a generic category page that doesn’t actually serve the searcher’s intent. Google explicitly discourages this kind of irrelevant redirect; it functions similarly to a soft 404 from Google’s perspective because the destination doesn’t answer the query that brought the user there, and it can also dilute the relevance of the page receiving all those irrelevant redirects.
The underlying mechanism worth understanding here is that redirects don’t transfer ranking signals unconditionally; Google’s systems evaluate whether the destination is a genuine continuation of the same content and intent before treating the redirect as a clean signal-passing event. A redirect from a discontinued blender model to the general “kitchen appliances” category page might pass some residual signal, but it won’t inherit the original page’s specific ranking position for long, because the query-to-content match has weakened considerably. Users who click through from a search result expecting a specific product and land on a broad category listing tend to bounce quickly, and that behavioral signal compounds the relevance mismatch already at play. This is why the guidance to redirect to “a genuinely relevant” destination is doing real work in the recommendation: relevance here means the destination should let the user complete something close to the same task, not merely share a parent category.
There’s also an important middle case worth naming explicitly: a discontinued product with a direct successor model. When a manufacturer replaces “Widget Pro 3” with “Widget Pro 4” that serves the same function for the same buyer, redirecting the old product URL straight to the new model’s page is usually the cleanest outcome, since search intent, content, and purchase path all line up closely. This is different from redirecting to a same-tier alternative from a different line or a loosely related accessory, which tends to underperform because the searcher’s specific intent (this exact model, at this price point, with these exact specs) isn’t actually satisfied by a same-category substitute, even though the redirect looks superficially reasonable from a site-architecture standpoint.
The comparison in practice
| Scenario | Status code | Indexing | Content treatment |
|---|---|---|---|
| Temporarily out of stock | 200 | Stays indexed | Keep page live, mark unavailable clearly (structured data availability field), show restock date if known, suggest similar in-stock alternatives without removing the original product content |
| Permanently discontinued, good replacement exists | 301 | Redirect target gets indexed, old URL retired | Redirect to the specific replacement product or a tightly relevant category/collection page, not a generic destination |
| Permanently discontinued, no good replacement | 404 or 410 | De-indexed over time | Serve a proper not-found response; a well-designed 404 page can still surface related in-stock products, but the status code itself should honestly reflect that this exact page is gone |
The distinction between 404 and 410 is worth a brief note: 410 (Gone) signals more explicitly that the removal is intentional and permanent, versus 404 (Not Found) which is more ambiguous about whether the absence is temporary. Google has said it treats both similarly in practice for crawling purposes over time, so this is a minor signal-clarity choice rather than a major ranking lever.
Handling the structured data layer
For out-of-stock products, Product structured data should reflect the current availability status (an “OutOfStock” value rather than removing the schema entirely), and if a restock date is known, communicating it in the visible page content helps both users and Google understand this is a temporary state. This is also where e-commerce platforms most often get this wrong by default: many platform configurations auto-remove a product from the sitemap or apply a noindex tag the moment inventory hits zero, which is precisely the behavior that erodes ranking equity for products the merchant fully intends to restock. Checking how your specific platform (Shopify, Magento, a custom stack) handles zero-inventory states by default is a necessary technical audit step, because the platform’s default behavior often contradicts the correct SEO handling described here.
Beyond the top-level availability field, a few adjacent structured-data details are worth getting right because they affect how the page is represented in rich results even while the product is unavailable. Price, currency, and offer validity fields should stay populated rather than blanked out, since a blank or clearly stale price field can trigger structured-data validation warnings and reduce eligibility for rich results even after the product restocks and availability flips back. If the product is unavailable indefinitely with genuinely no restock date, some sites use a “PreOrder” or “BackOrder” availability value where applicable instead of a flat “OutOfStock,” which more precisely reflects the actual purchasing state and can be more informative to both users and Google than a generic unavailable flag. None of these structured-data choices are ranking levers on their own, but they affect whether the page is correctly understood and correctly represented, which matters for click-through even when raw position doesn’t move.
It’s also worth addressing what happens on the category and collection pages that link to an out-of-stock product, since this is a layer merchants often overlook. If a category page’s internal logic automatically hides out-of-stock items from listings, that’s reasonable for user experience, but if the same logic also strips the internal link to the product page entirely rather than just suppressing its storefront visibility, it can slow Google’s ability to re-crawl that product page regularly, since internal links are a primary discovery and crawl-priority signal. A product that’s temporarily hidden from browsing but still linked from an XML sitemap and still reachable through at least some path on the site tends to get re-crawled and re-evaluated more reliably than one that becomes fully orphaned the moment inventory hits zero.
What this looks like operationally at scale
For a catalog with thousands of SKUs cycling in and out of stock regularly (seasonal goods, limited runs, supplier-dependent availability), the practical approach is building a rule into the product data pipeline itself rather than handling each SKU manually: a status field that distinguishes “temporarily unavailable, expected back” from “discontinued, no restock planned,” with each status driving a different automated behavior (keep live and indexed vs. trigger a redirect-mapping workflow). Products that sit in an ambiguous unavailable state for an extended period without a clear restock date are the hardest case; a reasonable practice is to set an internal threshold (for example, several months with no restock and no active reorder) after which the product gets reclassified as discontinued and moved into the redirect or retirement workflow, rather than leaving indexed pages for products that are functionally dead in an in-between state indefinitely.
Practical implication
Audit your platform’s default out-of-stock behavior first, since many CMS and e-commerce platforms silently do the wrong thing (auto-noindex, auto-removal from sitemap) without anyone deciding that on purpose. Build a genuine two-state classification (temporary vs. permanent) into your product data model rather than treating “out of stock” as a single undifferentiated state, and make sure whoever owns redirect mapping for discontinued products is redirecting to genuinely relevant destinations, not defaulting to the homepage or top-level category just to have somewhere to point the URL.
