A mistake that 301ed a lot of pages

[Joshmc]

Hey the developer made a mistake on my site where he removed a character from urls and it ended up 301ing about 20k pages of the 300k that are in googles index. This happened about 2 weeks ago and it has been indexed with the new urls but webmaster tools has 30k duplicate title tag errors, which is how I even realized this happened.

I was wondering, should I spend the time to un dev what happened back to the old links or should I just move forward with how it is since they 301.

Im asking this because my long tail seems to have suffered after goolge indexed the new pages but I want to know whether un 301ing them will even help it come back.

Thanks!

[tedster]

My preference would be to back things up to their original condition.

However, I don't understand how this created 30,000 duplicate titles. If there's a true 301 redirect, then googlebot shouldn't even get any text for the original URL. It sounds like there might be something else going on in addition to the problem you described.

[kellyman]

I Missed a Http:// from an internal link pointing to my homepage which resulted in Google seeing my home page redirected to itself

I lost 90% of traffic, i resolved the issue by backing up the original copy as stated by tedster and 301 the the page back to the original

24 hours later most of my traffic returned along with the PR i lost

However 1 year on i struggle with ranking the keywords i was targeting at the time and all top 5 in Google so never really fully recovered

But in answer to your question put things back to how they were

[Joshmc]

@tedster, yea that is a little weird I checked the old urls via the redirect checker and they are showing valid 301 moved permanently codes, however they show as duplicate title tags in webmaster tools which is strange...

[aakk9999]

Tedster is right that it is not possible to get 30,000 duplicate titles from the description of what happened. If a character of URL was removed, which in turn caused (lets say) that "old URL" redirect to a new URL with removed character, then Google would never report duplicate title.

The only way of you having a duplicate titles would be the influx of URLs (with removed character?) returning the same content as the old URLs. In this case, even if you set up redirect from "old" URL to these "new" URL (with removed character), Google will not be able to see 301 redirect until Google re-crawls the old URLs. This would result in duplicate titles showing in WMT for a period of time, until all "old" URLs are re-crawled.

So I think you really need to ask your developer what did he exactly do and in what order, especially whether he set up 301 afterwards in order to try to rectify the missing character errors?

It would be wise to know exactly what happened so that you do not end up sending 301 back and forth between "old" and "new" URLs in order to rectify the problem.

<added> What URLs are currently linked from within the site - old URLs or new URLs with missing character? I.e. do you now have lots of internal 301 redirects (site shows old URLS that do 301) or does the internal links show new URLs with missing character?</added>

[Joshmc]

@aakk9999 the site does currently link to the new versions of the urls without the special character. The way oursite is set up you can have example.com/productx/cat/124 and no matter if you enter something like example.com/prx/cat/124 it will auto redirect back to example.com/productx/cat/124. This is the way the system was set up so the urls 301s should have been instant when the special characters were removed.

It could be a case of Google needing to recrawl the pages in order to get this updated, I will have to let you know as it goes on, often it takes a long time for the duplicate title tag error to remove itself from webmaster tools though.

As I have been pouring over the answers and what to do, I think i am going to have to bite the bullet and leave the new urls in place. They are almost all indexed now with the new url and im afraid of what will happen if i 301 them back again, seems like Google would hate this back and forth. Also we are a small company so the dev time to unwrite the removal of special characters (and hoping it creates the old urls again) seems like it may take to much of the developers time away from other projects he is working on for us.

Let me know if you think that i absolutely should change them back and then I will atleast look into it further. Thanks for all the great help on this question @tedster, @kellyman and @aakk9999

[seoskunk]

I would reverse the mistake and 301 the new urls

[aakk9999]

OK, I think I got it.

So you had this: example.com/productx/cat/124 as internal links and anything else with "124" did 301 to example.com/productx/cat/124 (which is the right way to do it)

The example.com/productx/cat/124 URL format was indexed.

Now the character was removed and it affected both, your internal links AND your redirect.

So, lets say that for example, the character removed was x from "productx". Hence now internal links are in the format of example.com/product/cat/124 and also any other format of URL with "124" now redirect to this new version of URL. Consequently, all your "old" links would automatically redirect to the new URL format.

As you say, when the mistake was made, all internal links also took the format of the "new URLs".

Here is how Google saw it:

- It knows of nnn links in the "old format" and these are the links it has found on the site before.
- Suddenly, it recrawls the site and it finds zzz "new" links and the old links are not there (as your old links are now replaced with "new format"). Even though it did not find "old" links any more, it will hang onto them in index as it knows they exist(ed) - until it recrawls them.
- In most cases, Google is quicker to crawl new links than to re-crawl existing. It is likely that looking at crawl activity in WMT you will find a spike in crawl which will correspond with when Google hit the site and suddenly found many "new" links.
- So, Google found many links with a new link format, but has not re-crawled old link format yet (hence it is not yet seeing 301 redirect from old to new).

This is the cause of your "duplicate titles" in WMT.

Tedster and seoskunk both think you should revert back as soon as possible. I think the same.

I would bite the bullet and revert. Then wait until Google sort itself out. If this is one-off mistake (i.e. the site was stable with regards to links before), and providing you get "reversion" correct, you should see the full recovery within certain time (month to three months usually, depending on crawl rate etc.).