Wrong canonical URL was specified. How to refresh the index now?

[Krupesh]

Wrong canonical URL was applied to thousands of pages of a client website, pointing them all to a single non-existing URL. Now Google has de-indexed most of those pages. We have fixed the problem now, but do we get Search engines crawl those pages again and start showing in Search results?

I understand that a slow recovery is possible if we don't do anything. Was wondering if we can fast track the recovery... Any pointers?

Thanks

[Highland]

Canonical isn't like a 301, where the page is eventually dropped. Canonical is a hint the page gives to what should win the duplicate content race. It doesn't mean you won't be crawled again, but it might take some time. The key factor here is page popularity. The more popular a page, the faster it gets crawled.

Have you considered a social campaign for the pages in question?

[LynnPatchett]

Hi,

You could try fetching the page (homepage I guess unless these pages were all part of a sub section of the site) in webmaster tools which should help speed the process, see this page for details. And of course if not already done make sure you have a valid sitemap in GWT which has all the relevant urls in it. A bit of patience and they should come back.

[Dr-Pete]

Yeah, this is a good starting point. Create a sitemap in GWT with just these pages (it's easier to monitor that way), and re-fetch any specific pages that are critical. You can also build internal links (even temporarily) to kick the crawlers or try promoting some pages via Google+. There's no foolproof method, though - just nudges.

This assumes, of course, that you've corrected the canonical tags. If there shouldn't have been a canonical tag at all, then I'd recommend adding a self-referencing canonical (i.e. one pointing to the page itself). A new canonical tag seems to overwrite an old one better than just removing, at least that's my anecdotal observation.