Deleted Pages and SEO: When a Redirect Has Nowhere Useful to Go

Deleting a page does not mean it should redirect to the homepage or nearest category. Use a relevance test to choose between a replacement, temporary withdrawal, improvement or clean removal.

When a team removes an old page, one awkward question follows: what should happen to the URL?

The usual answer is to redirect it to the homepage or the nearest broad category. Leaving a broken URL feels untidy, while a redirect seems like the responsible option. But a redirect is not simply a way to make an old address disappear. It tells visitors and search engines that another URL has meaningfully replaced what used to be there.

That distinction matters. A genuinely merged service may have an obvious replacement. An expired campaign may not. A discontinued product might have a successor, but only if that successor still meets the original buyer’s need. An outdated article may be better improved than removed.

This article sets out a practical test: does the proposed destination satisfy substantially the same visitor need, preserve the original page’s important meaning and provide a sensible next step?

A redirect should represent a real replacement

In HTTP terms, a 301 response indicates that a resource has been assigned a new permanent URI, or web address. The protocol establishes what the status means, but it does not decide whether a particular destination is relevant to a visitor. See RFC 9110 for the technical definition.

Google’s documentation similarly recommends a permanent redirect when content has moved or has a clear replacement. When removed content has no similar replacement, Google recommends returning a proper 404 or 410. Its guidance uses terms such as “clear replacement” and “similar content”, but does not publish a formal similarity score or numeric threshold.

That leaves a useful practical question for marketers and SEO teams:

Would a visitor arriving at the new URL feel that they had reached what they were looking for, or would they have to start the search again?

A practical relevance test has three parts:

• Same need: does the destination address substantially the same question, task or commercial need?
• Same important meaning: does it preserve the key subject, promise, product purpose or service context of the original page?
• Sensible next step: can the visitor act, learn or continue without having to reinterpret why they were sent there?

If all three are true, a redirect is usually defensible. If the pages are only vaguely related, the redirect is probably being used because it is convenient rather than accurate. This is an applied decision model, not a Google ranking formula.

Why the homepage is usually the wrong answer

The homepage is easy to select because it is stable, prominent and unlikely to disappear. It is also a poor substitute for most specific pages.

Someone following a link to a discontinued camera, a 2023 conference campaign or a specialist service does not arrive on the homepage with their original need solved. They arrive somewhere else and must begin navigating from scratch. That can create a confusing user journey and make the original page’s lifecycle harder to interpret in analytics and technical diagnosis.

Google warns that redirecting many unrelated old URLs to one destination, such as the homepage, may confuse users and may be treated as a soft 404. A soft 404 is a page that technically responds successfully but behaves, from a search perspective, as though the requested content is unavailable. This does not mean every homepage redirect causes a specific ranking loss. It means that a large collection of unrelated redirects is a warning sign, rather than a tidy clean-up strategy.

The nearest category has a similar trap. A product URL can be sent to its category because the category is topically related. But “related” is not the same as “replacement”. A visitor looking for a specific waterproof hiking jacket may benefit from a category page if the product was a thin item within an active range and the category offers obvious alternatives. Someone looking for a discontinued specialist lens may need compatibility information, a successor model or a support explanation that the category page does not provide.

Treat the homepage and broad categories as warning cases. They can sometimes be appropriate, but proximity alone is not enough.

Different page lifecycles need different decisions

Deleted pages often get handled by one rule because the CMS makes one rule easy to apply. A better starting point is to ask what happened to the thing the page represented.

Expired campaigns

An old campaign page usually represented a time-bound offer, event or promotion. Once the offer has ended, there may be no equivalent destination.

Suppose a retailer has a URL for a “January 2025 home office sale”. Redirecting it to the homepage says very little. Redirecting it to the current sale category may also be misleading if the products, terms and customer expectations are different. The old page promised a particular promotion; the new page does not necessarily replace that promise.

Possible outcomes include:

• retain and update the page if it has useful historical, customer-service or recurring-campaign value;
• redirect it to a genuinely equivalent successor campaign if one exists and the visitor’s likely purpose remains the same;
• remove it cleanly when the campaign has ended and there is no useful successor.

A 410 has a specific HTTP meaning for a resource that is no longer available and is likely to remain unavailable. A 404 indicates that the server cannot find a current representation without stating whether the absence is temporary or permanent. See RFC 9110 for these protocol definitions.

Google has stated that it treats 404 and 410 the same for Google Search, so 410 should not be presented as an automatic SEO improvement. That statement is specific to Google Search and does not remove the operational distinction between the two HTTP statuses.

Discontinued products

A discontinued product can have a valid replacement, but the replacement must be more than the next item in the catalogue.

Imagine a manufacturer removes a 2021 smart thermostat. Its new model may be a sensible destination if it performs the same core job, works with the same expected systems and answers the buyer’s likely next question. If the new model uses a different installation method or is aimed at a different type of property, a direct redirect may leave important expectations unanswered.

A useful check is to compare:

• the original product’s main purpose;
• compatibility and specification expectations;
• the type of buyer and use case;
• support, warranty or replacement questions;
• whether the proposed successor is actually available.

If the successor preserves the original product’s purpose, meaning and sensible next step, a redirect may be appropriate. If not, the page may need a short discontinuation explanation, a support destination or a clean removal outcome. An unavailable product is not automatically a deleted product, which is why out-of-stock product pages need their own lifecycle decision.

Merged services

Merged services are often the strongest case for a redirect. If two service pages are combined into one offer, and the new page clearly covers the old service’s subject, deliverables and customer need, it may be a genuine replacement.

For example, a consultancy might retire separate pages for “market research” and “customer insight” after creating one service that deliberately combines both. A redirect from either old page to the new service page could make sense if the new page still explains the relevant work and gives that audience a clear route to enquire.

Changing the label is not enough. If the new page only describes customer insight and drops the market-research work that attracted the old audience, it is related rather than equivalent. The redirect may conceal a content gap instead of solving one.

When a consolidation is genuine, map the old page’s important promises against the new page. If the new page cannot support those promises, improve it first or choose a different outcome. The guide to content consolidation and deciding whether to merge, redirect or keep a page covers the broader consolidation question.

Outdated articles

An outdated article is not necessarily disposable. Its subject may still matter, while the facts, examples or recommendations need updating.

Consider an article explaining how small businesses should prepare for a tax-year change. If the underlying topic remains useful, updating the article on its existing URL may preserve its history, references and search intent more accurately than redirecting it to a generic finance hub. The page has not been replaced; it has become stale.

Deletion becomes more reasonable when the subject is no longer useful, the information cannot safely be corrected or the article’s original purpose has genuinely been absorbed by a substantially different resource. Even then, the proposed destination should answer the same question. A general “resources” page is not automatically a replacement for a detailed guide.

Low recent traffic is not enough to justify deletion. Backlinks, internal links, referral traffic, customer-support use, historical importance and the possibility of a useful update all add context. Traffic is evidence, not a verdict.

A practical decision flow

Before choosing a destination, classify what has happened to the original page.

• There is a genuine replacement: redirect when another URL satisfies substantially the same visitor need, preserves the important meaning and provides a sensible next step.
• The destination is related but not equivalent: do not redirect merely because it shares a category, keyword or navigation path. Consider improving the destination, retaining a useful explanation or removing the old URL cleanly.
• There is no useful destination: a 404 or 410 can be the clearer outcome. A 404 is not automatically harmful to the rest of a site, according to Google’s documentation on 404 errors. The quality of the error experience and the handling of internal links still matter in practice.
• The withdrawal is temporary: do not treat a temporarily unavailable page as permanently deleted. If visitors are sent elsewhere for a temporary period, the alternative should itself be genuinely temporary and appropriate to the original need. The relevant HTTP semantics are described in RFC 9110 and Google’s redirect documentation.
• The content is still useful but outdated: improve or update it instead of deleting it. The existing URL may still be the best home for the subject.

This flow is deliberately cautious. It does not claim that every redirect passes value, that every 404 is harmful or that a 410 is always preferable. It asks what the URL represented and whether the proposed outcome is honest about that history.

What to check before deleting the URL

Make the decision with evidence rather than the CMS’s default rule. Before deletion, check:

• Backlinks: are other sites referring to the page, and would a genuine replacement help those visitors?
• Organic entrances: does the URL still attract searches, including specific long-tail queries?
• Internal links: do important pages, navigation elements or tools still send visitors there?
• Referral traffic: are partners, campaigns or customer-service teams using the URL?
• Customer expectations: what would someone following a saved bookmark, support link or partner reference expect to find?
• Changed form: does the original subject still exist as a new product, service, offer or question?

These checks do not produce the answer mechanically. They reveal the consequences of each option. A URL with little current traffic may still deserve an update. A heavily linked URL may deserve a redirect only if a relevant successor exists. A page with many internal links may expose an architecture problem rather than a redirect opportunity. Google’s site-move guidance provides relevant context for checking URLs, links and redirects, but it does not turn these checks into an automatic decision rule.

Why this matters at scale

On a small site, these decisions can be made URL by URL. Someone can read the old page, inspect the proposed destination and ask whether the visitor’s need is genuinely preserved.

Large sites need a repeatable policy. Otherwise, an automated deletion rule can turn thousands of expired campaigns, discontinued products and old articles into redirects to one homepage or broad category. That may look efficient in a spreadsheet while sending real visitors to irrelevant destinations.

A scalable process should record the page lifecycle, proposed outcome, destination and reason. QA should test that redirects point to relevant replacements, temporary withdrawals are not treated as permanent removals and cleanly deleted pages are not being funnelled into a catch-all URL.

Automation is useful for collecting backlinks, entrances, internal links and referral data. It is less reliable at judging whether a specialist product and its successor preserve the same customer need. That judgement is where human review, clear criteria and careful sampling earn their keep.

Conclusion: remove the page, not the meaning

A redirect is for a meaningful replacement, not for any URL that happens to be nearby.

Ask whether the destination meets the same visitor need, preserves the original page’s important meaning and offers a sensible next step. A merged service may pass that test. An expired campaign may not. A discontinued product may need a successor or a support explanation. An outdated article may deserve an update rather than deletion.

When there is no honest replacement, a clean 404 or 410 is often clearer than sending visitors somewhere unrelated. When the decision affects thousands of URLs, the same relevance test can become a documented lifecycle policy: classify what changed, assess the proposed destination and make the reason for the outcome explicit.