Pages with an end date
Most sites have pages that are meant to last. Listings, jobs, event and course sites have the opposite: pages that are created in bulk, live for days or weeks and then stop being true. A flat is let, a vacancy is filled, a concert takes place, a course run starts, a sale finishes. The URL, the links to it and anything search engines have stored about it remain.
What happens next is usually decided by accident. The template keeps rendering with the details stripped out, leaving a page that returns success but shows nothing, which Google reports as a Soft 404A page that returns a success status code but shows an error, empty or not-found message.. Or every expired URL is redirected to the home page. Or the page stays exactly as it was, so searchers find a job they cannot apply for. Each of those fails the person who clicks, and on a large site each one multiplies across thousands of addresses.
Pages with an end date differ from out-of-stock products, which usually come back and keep their value; product availability states are covered in the category and product page guide. They also differ from the soft not-found symptom itself, which the soft 404 troubleshooting page diagnoses once it appears. This guide is the policy that prevents it.
The options when a page expires
There are four honest options, and one dishonest one to avoid.
- Keep the page, marked as expired, with alternatives. The URL stays indexableWhether a URL is eligible to be added to a search engine's index: reachable, returning success, not excluded by noindex and not canonicalised elsewhere., the content stays, and the page says clearly that the item has ended and offers what a visitor can do instead. Fits pages that still have value: a past event with a recap, a sold property whose details and photographs remain useful to people researching the area, a course that runs again.
- Keep the page for visitors, out of the index. The page stays for people who follow old links, with a noindexA robots directive telling search engines not to show a page in their results. so it leaves search results. Google's noindex documentation explains that the page is dropped when it is next crawled and the rule is seen. Google's crawl budget guide adds that noindexed pages are still requested, so on a very large site this option still costs crawling.
- Redirect to a true equivalent. A permanent redirect to the page that replaces it: the same vacancy reposted at a new address, the next edition of a recurring event, the successor to an offer. Google's redirects documentation describes a permanent redirect as a strong signal that the target should be canonical; that is only honest when the target is the same thing.
- Remove the page with a not-found or gone response. The URL returns 404 or 410, internal links are removed and the URL leaves the sitemap. Google's crawl budget guide describes this as the right response for permanently removed pages and a strong signal not to crawl the URL again.
The dishonest option is the success response for an empty page, or the redirect of everything to the home page or a broad category. Both leave the visitor with nothing they asked for, and both are reported as soft errors.
Choosing by page type
The page type sets the default; the page's own value decides exceptions. A page with external links, steady search visits or content worth reading after the end date can justify keeping when its type says remove.
| Page type | Value after expiry | Default option | Exception |
|---|---|---|---|
| Rental listing, let | Low; the property may return | Keep briefly with a "let" notice and similar homes, then remove | Same URL revived if the property is relisted |
| Property for sale, sold | Some; people research sold homes in an area | Keep with "sold" notice, details and area links | Remove if details must be withdrawn |
| Job posting, filled or closed | Low | Remove, or keep briefly with similar vacancies | Redirect if the same role is reposted at a new URL |
| Classified ad | Low | Remove | None usually |
| One-off event, past | Some, if a recap is added | Keep with recap and upcoming events | Remove if the page was only a ticket form |
| Recurring event, past edition | Low for the edition; high for the series | Redirect to the evergreen series page, or keep as an archived edition linked from it | Keep editions with rich recaps |
| Course run, started | Low for the run; high for the course | One evergreen course page with dates as content, not separate URLs | None |
| Time-limited offer, ended | Low for the offer | Redirect to the offer's stable page, or remove | Keep if it returns each year at the same address |
A branch's location page is a different case, because a moved branch keeps its page and a closed one usually points visitors to the nearest open site; the tutorial on moving or closing a location applies these options to the page alongside the profile, the listings and the ads.
Keeping an expired page useful
A page that stays live after its item has ended must say so at once and offer a next step. Practice that works:
- A clear notice at the top: let, sold, filled, closed, ended, with the date. Never leave a visitor to discover the problem at the application form or the checkout.
- Alternatives that match the original intent. Similar homes in the same area and price band, open vacancies in the same role and location, the next date of the event or course, the current offer. Generated from the item's own attributes, not a generic carousel.
- The content that is still true. Photographs and details of a sold property, the recap and speakers of a past event. A page stripped to a heading looks empty because it is.
- No dead actions. Remove or disable apply, book and buy buttons, and update any markup that still describes the item as available, so the page and its structured data agree.
- A link to the evergreen parent, such as the area page, the job category or the event series.
Whether a kept page stays indexable is the question from the table: indexable if a searcher would be glad to land on it, noindexed or removed if not.
Status codes, sitemaps and links at expiry
Expiry changes more than the page. At the moment an item ends:
- The status code follows the decision. Success for kept pages, a permanent 301 RedirectA permanent server redirect telling browsers and search engines a URL has moved for good. for redirected ones, not-found or gone for removed ones. Google's crawl budget guide asks sites to eliminate soft not-found errors, noting that such pages continue to be crawled.
- The XML SitemapA file listing the URLs a site wants search engines to discover and crawl. changes the same day. Removed and redirected URLs leave it; kept pages that are noindexed leave it too. The crawl budget guide asks for sitemaps kept up to date, which on a listings site means regenerated as items change state.
- Internal links are removed from listing pages, search results and related-item blocks. An Internal LinkA link from one page to another page on the same website. to a removed page sends visitors and crawlers to a dead end; links to kept pages move from the current listings to an archive or the parent page.
- Redirect chains are avoided. If an expired URL redirects to a page that later expires too, update the first redirect to the final destination.
Search Console's Page Indexing report then shows the result: removed pages under not found, noindexed pages as excluded by the noindex tag, redirected URLs as pages with redirects, and any soft errors that remain. Not-found entries for pages removed on purpose are expected; soft errors and not-found entries for live pages are the ones to investigate. The crawling and indexing guide explains how each response is treated, and the page on not-found errors in Search Console separates expected removals from leaking links, lost redirects and live pages answering not found by mistake.
Evergreen parents for recurring pages
Many things with end dates come back: an annual conference, a course that runs every term, a seasonal sale, a property that is relisted, a role that is hired for repeatedly. Giving each occurrence a new URL means every edition starts from nothing and expires with whatever links and rankings it earned. An evergreen parent keeps them.
- Events: one stable page for the series that always describes the next edition, with past editions as archived children linked from it, or redirected to it.
- Courses: one page per course, with run dates and booking as content on the page rather than as separate URLs.
- Offers: one stable address for a recurring promotion, updated each time it runs and kept between runs with a note on when it returns.
- Listings: area, property type and category pages that are always live and always list current stock, so the searches that individual listings briefly rank for are served permanently by the parent.
- Jobs: role and location pages describing the work, with current vacancies listed on them.
The parent is where links and rankings accumulate, and the item pages become short-lived children that can expire without loss.
Running expiry at scale
On a site where hundreds of items change state each day, expiry must be a rule in code, not an editorial task. Store the end date and the state on every item. Map each state to its option, its status codeThe numeric code a web server returns with every response, saying whether the request succeeded, was redirected or failed and why., its sitemap behaviour and its notice, in one configuration the templates read. Run a scheduled job that applies state changes, regenerates sitemaps and updates listing pages, and log what it did.
Google's crawl budget guide is written for large sites and for sites whose content changes very quickly, which describes many listings and jobs sites, and its advice is the same as this policy: consolidate duplicates, return honest status codes for removed pages, avoid soft errors, keep sitemaps current and avoid long redirect chains. Check the result through the same lens: the soft error and not-found rows of the Page Indexing report after each change to the rules, crawler requests to expired URLs in the server logs, and the indexed count of live items against the number that exist.
Snugwillow Lettings writes an expiry rule
Snugwillow Lettings, an invented letting agency, publishes about 1,500 rental listings a year, most let within three weeks. Let listings kept their URL with the photographs and description removed, and the Page Indexing report showed 940 soft errors, while only 210 of its 330 live listings were indexed. The agency sets one rule per state. Let: the page shows a "let" notice, the original details and up to six similar homes for 14 days, leaves the sitemap and listing pages at once, and then returns a gone response. Withdrawn: gone at once. Relisted: the same URL returns to the sitemap with fresh details. It also builds evergreen pages for each of its 12 areas, listing current homes and linking to the area guides. Three months later, soft errors are down to 35, 312 of 340 live listings are indexed, and the area pages receive more organic visits than individual listings did before.
Common mistakes
- Stripping the page but returning success. The commonest cause of soft errors on listings sites.
- Redirecting every expired page to the home page. Visitors get nothing they asked for, and the redirects are treated as errors.
- Leaving expired items in sitemaps and listing pages. Crawlers keep requesting what the site has already given up.
- Noindex as the default for millions of pages. It removes them from results but not from crawling.
- A new URL for every occurrence of a recurring thing. Each edition starts from nothing and loses what the last one earned.
- Deciding at expiry instead of before. Without a written rule, each template does something different.
- Dead actions on kept pages. An apply or buy button on an ended item frustrates the visitor and misleads.
