How to restore a website from the Wayback Machine

If a website has gone (the hosting lapsed, the backups lived on the same server, the developer stopped answering), public web archives such as the Internet Archive's Wayback Machine are often the only place its pages still exist. You can rebuild a good deal from them. What comes back is the site as a crawler saw it from the outside: the page text, many of the images, some of the stylesheets. What does not come back is the database, the content management system, the plugins, or anything that only worked by talking to the original server.

There is also a question most restoration guides skip: whose content is it? The Internet Archive's own help pages say its terms "do not cover backups for the general public", but that you may use the Wayback Machine "to locate and access archived versions of a site to which you own the rights". That last phrase decides much of what follows.

Two quite different people search for this. One has lost their own site. The other has bought or caught a domain and found, in the archive, a website that belonged to somebody else. The mechanics of recovering pages are the same for both. What each may do with those pages is not.

The table is a starting point, not a legal opinion. This guide is general information about UK law and practice, not legal advice; if a lot rides on reusing particular content, take advice on your own facts.

An archive copy is a last resort, because it records what a visitor saw rather than what the site was made of. A real backup brings back the lot, including the parts no archive can hold. Before you start, check:

• your hosting account, including any old or cancelled plans the host may still hold data for;
• exports from the CMS or website builder;
• the developer or agency that built or maintained the site;
• old email threads, shared drives and laptops, for copies of text, logos and photographs.

Do not expect the Internet Archive to fill the gap on request. Its help page is direct about it: "We can no longer offer the service to pack up sites that have been lost", and "We can't guarantee that your site has been or will be archived."

This is the order we would work in. The early steps establish what exists before anyone promises a restored site. The later ones stop the rebuild from breaking the addresses that other sites, search engines and customers still use.

Steps two and three are where most of the effort goes. If you are comfortable with an API, how Wayback Machine restoration works shows how to build the inventory from the CDX index rather than by clicking through the calendar, and how to spot parking-page captures in the data. On step six, assume outbound links have decayed until you have checked them. Pew Research Center's 2024 study of link rot put a number on it: "38% of webpages that existed in 2013 are no longer accessible a decade later".

Every archived site has gaps. The Internet Archive's help pages explain most of them, and it is worth knowing the causes before promising anyone a complete restore.

The last row catches people out. When you browse an incomplete capture, the Wayback Machine "will grab the closest available date to the one you are in for the links that are missing". If it holds nothing for a link, it "will look for the link on the live web and grab it if available". So a page that looks right in the browser can be a 2014 layout wearing a 2018 stylesheet, with an image served by whatever lives at that address today. Check the date of every asset you keep, not only the page's.

Dropped domains add a further trap: the archive's last good view of the site can be old. Across 7,310 recently dropped .uk domains with archived homepage history, DropOps found a median gap of 519 days between the last working homepage capture and the drop. For 1,230 of the 7,001 domains with a year-by-year record, the final archived year held no working homepage at all, only redirects or errors. The sample leans towards domains with strong link evidence, so read these as a description of that sample rather than a rate for every domain; what can actually be recovered from an archived website sets out the full study.

Copyright in the UK is automatic. GOV.UK lists "web content" and photography among the works protected from the moment they are created, with no register and no need to mark them. For written and artistic work it lasts "70 years after the author's death", so an old website's text and photographs are almost always still in copyright.

Who owns them is set by the Copyright, Designs and Patents Act 1988. Section 11 makes the author the first owner, except that where a work "is made by an employee in the course of his employment, his employer is the first owner of any copyright in the work subject to any agreement to the contrary". Freelancers are the other way round: GOV.UK says that if you are self-employed, "you usually own the intellectual property even if your work was commissioned by someone else - unless your contract with them gives them the rights". A business restoring its own site may therefore own the words its staff wrote but not the photographs a freelancer took or the design an agency produced, unless the paperwork says otherwise.

The Internet Archive does not settle any of this for you. Its Rights help page says it "does not make guarantees as to the copyright status of items", and its terms of use ask users to certify that their use "will be limited to noninfringing or fair use under copyright law". Fair use is a US legal concept; UK law has its own, narrower "fair dealing" exceptions. Presence in the archive tells you a page existed. It does not tell you that you may copy it.

Three more things to check

• Database right. If the old site carried a directory, listings or a product catalogue, the contents may be protected in their own right. GOV.UK explains that database right protects "the contents of a database" where there was "a substantial investment in obtaining, verifying or presenting the data". The regulations give it fifteen years from the end of the year the database was completed, and a substantial change can start a new term.
• Personal data. Testimonials, staff pages and named customer photos are about identifiable people. The ICO notes that information about sole traders, employees and company directors "may constitute personal data", and its guidance says personal information should be reused for a new purpose only if that is "compatible with your original purpose", and that any new purpose needs a lawful basis of its own.
• Names and trade marks. Rebuilding a former business's site under its name and look can suggest that the business is still trading, or that you are it. That is the territory of registered trade marks and of passing off, which HMRC's manual sums up as "the classic trinity of reputation (or goodwill), misrepresentation and damage".

Whatever you decide, keep the evidence: the contract or invoice that assigned the rights, the email granting permission, the date you checked a stock licence. A written record answers a later question in a way memory cannot.

Catching a dropped .uk name, or buying an expired one, gives you the address and whatever value it still carries in links and visitors. It does not give you the previous owner's website. The defensible use of the archive here is as a map rather than a source: which pages existed, which addresses other sites linked to, what the site was about and when it was active.

Google's position cuts both ways. Its spam policies define expired domain abuse as an expired domain "purchased and repurposed primarily to manipulate search rankings by hosting content that provides little to no value to users". The March 2024 announcement that introduced the policy also says: "It's fine to use an old domain name for a new, original site that's designed to serve people first." Google's site-move guidance adds a practical step for any recently bought domain: check Search Console for manual actions and for URL removals "left over from the previous owner".

For an acquired domain, then, the realistic choices are to repurpose it for something related or to start fresh, while still deciding what happens to every old address people and other sites point at. Restore, modernise, repurpose or start fresh? works through that decision.

DropOps Sites is a website builder for domains you already own. It reads a domain's old addresses and a sample of its old pages from public web archives to learn what the site covered, how it was organised and when it was active. Bringing the previous site back as closely as possible (the "Restore original" approach) is offered only once the owner has stated that the content is theirs to restore, because they own it, represent the owner or have permission. We record that statement; we do not make a legal judgement on it. Recovered details that date quickly, such as phone numbers, prices, addresses, testimonials and opening hours, stay on hold until the owner confirms, replaces or removes them. Scripts, trackers and embeds are stripped, nothing is hotlinked from the archive, and nothing goes live until the owner publishes it.

A restored site that changes every URL throws away the links other sites made to it and the bookmarks its visitors kept. The simplest protection is to put important pages back at exactly the addresses they had. Where that is not possible, Google recommends server-side permanent redirects, "such as 301 and 308", from each old URL to its new equivalent, advises against chains, and warns against redirecting "many old URLs to one irrelevant single URL destination, such as the home page", which "might be treated as a soft 404 error".

For addresses with no sensible new home, a 404 or 410 is the honest answer. Google says all 4xx errors except 429 "are treated the same", so choose between them on meaning, not on any expected search effect. Keep the redirects you do set up: Google's site-move page says "generally at least 1 year", and suggests considering keeping them indefinitely for users.

Finding every old address is a job in itself when the site no longer exists. How to recover old URLs and build a redirect map covers it, and how to rebuild an old website without losing its SEO covers the wider rebuild. The order matters more than the tools: establish what survives, settle what you may reuse, and only then decide what each old address should do.