Welcome to WebmasterWorld Guest from 220.127.116.11
. Will the backlink continue to show up as such in the SERPs?
. Does the (no longer existing) destination page stay in the index as URL only?
. Do said backlinks continue to carry any weight, particularly if there are links off a custom 404 page to the rest of the site?
Along the same lines; in a scenario where a site was rebuilt from the ground up with no attempt to map old URLs onto the new ones, would there be any *spidering* benefit in having a site map incorporated into a 404 page? Could the deep linking potentially speed up the spidering of the new pages, or would Googlebot stop when it hit the 404?
I'd assumed that this was because they hadn't mapped the old URLs to the new, but I've got no firsthand experience of the consequences of doing this because ... well ... it's a spectacularly daft thing to be doing in the first place.
As such, I'm just trying to get get clarification on what I *think* I know in regards to Google's handling of 404s. It's a quest for knowledge thing rather than a 'I'm about to do something stupid' thing. So no need to worry. :)
Use a mod rewrite (if using Apache or else the equivalent on your web server) to return a specific page whenever a 404 occurs. In effect, you will be presenting a specific page that will look like a 404 page to the average user (with sitemap), but will actually return a 200.
These pages will trigger the Google duplicate content filter, but pagerank should still be dispersed accordingly.
Besides, what if jetboy decides to link to a few hundred non-existent files on mrMister's site? That'll surely get you filtered. Do you really want to leave yourself open to that kind of attack?
mrMister, running the gauntlet of G's dupe content filter intentionally ... not convinced this is a reason you'd want to do this. A filtered site with a high pagerank is still a filtered site.
The only pages that will be filtered would be the 'fake' 404 pages. A 404 wouldn't be listed in Google's results anyway.
Although, I guess it's unlikely that Google will give any weight to links on a duplicate page.
OK, new idea....
Perhaps substituting the 404 error of the server to a 301 to the home page would work. Anyone have any experience of this?