Say your host went bust, or an old developer deleted the account, and the site you built is gone. A quick search gives you a dozen people offering to “restore” it. They are not all offering the same thing. These questions separate them.
Most of them can be answered before any money changes hands. If a provider dodges them, that is an answer too. Read the list once, then ask them in order.
1. Where will the content come from?
If the source is the Internet Archive’s Wayback Machine, the result is capped by what was captured. Its FAQ says broken images occur when the images are not available on its servers, and that simple HTML is the easiest to archive (Internet Archive Help Center). Ask the provider to show you what the archive holds for your domain before you commit. The free scan here does that.
The archive is not the only possible source. If you still have a database backup, a hosting snapshot, or files on an old drive, say so, because those are usually more complete than anything the crawler saved. The archive is the fallback when nothing else survives, not the first choice. A provider who assumes the archive without asking about your backups has skipped a step.
2. Which snapshot is used for each page?
“Latest” is not always best. A later capture may be a parked page or a different owner’s content. Ask how snapshots are chosen, and whether anyone checks the result.
This matters most for domains that lapsed and were re-registered. After a domain expires, the crawler often keeps capturing whatever the new owner parks there, so the newest date can be an advert page, not your site. The right snapshot is the most complete one from while you still owned the site, which is rarely the most recent. Ask how a provider tells those apart.
3. What happens to the old URLs?
Pages that keep their address keep their inbound links. The rest need a redirect to the closest page. Ask for the redirect map. The site migration checklist explains why.
A redirect map is just a list: each old address and where it now points. You want it in writing so you can check it, and so you can hand it to whoever hosts the new site. Without one, links from other sites and old bookmarks land on error pages, and that is the easiest part of a restore to get wrong.
4. What is left out, and will you be told?
Spam, thin and duplicate pages should be dropped, and the list should be given to you so you can disagree.
Old sites collect junk: hacked pages, auto-generated posts, broken drafts. Clearing them is sensible, but it is your site, so you should see what was removed and be able to put anything back. “Trust us, it was junk” is not good enough. Ask for the dropped list as plain text or a spreadsheet.
5. What do you get at the end?
An editable WordPress site, clean HTML, or only files? Ask for the format in writing. Make sure you receive the finished result, not a tool to run yourself.
There is a real difference between being handed a working, editable site and being handed a folder of files and a download script. Both are valid, but they are not the same purchase. If you are not technical, “a set of files” may mean you still have to hire someone to put them online. Confirm who does that last step.
6. Who checks the work?
My agents do the restores here; an LLM checks every page and a human reviews the job before delivery. Ask any provider, including me, what their equivalent step is. An agent can be wrong, like any colleague.
The question behind the question is whether a person ever looks at the result before it reaches you. Automated restores are fast, but they make confident mistakes: a parked page that looks plausible, a stylesheet from the wrong year. Someone checking the output against the original catches those. Ask whether that check happens, and whether you can review a sample before sign-off.
7. What are the rights and limits?
You should be asked to confirm you have the right to the content. See the terms and the DMCA policy. Be wary of anyone who promises a ranking outcome: nobody can.
A restored site is the content you once had, back online. It is not a guarantee that search engines will treat it the way they once did, and no honest provider will tell you otherwise. Restoring is a starting point. What happens afterwards is ordinary website work, and it is on you, not on the restore.
8. Do you need to hire at all?
For a small site and a free afternoon, the open-source downloader may be enough. The hire makes sense when the site is large, spans several eras, or has spam in its history. Compare plans or use the contact page.
Try the free route first on a few pages. If it works and you have the time, keep going. If it stalls, or the pages come back broken and you cannot tell which snapshot is right, that is the point where paying for it saves more than it costs.
A good provider answers the boring questions first.