Severe rank drop due to overwritten robots.txt

[BrettKrasnove]

Hi,

Last week we made a change to drupal core for an update to our website. We accidentally overwrote our good robots.txt that blocked hundreds of pages with the default drupal robots.txt. Several hours after that happened (and we didn't catch the mistake) our rankings dropped from mostly first, second place in Google organic to bottom and mid first page.

Basically I believe we flooded the index with very low quality pages at once and threw a red flag and we got de-ranked.

We have since fixed the robots.txt and have been re-crawled but have not seen a return in rank.

Would this be a safe assumption of what happened? I haven't seen any other sites getting hit in the retail vertical yet in regards to any Panda 2.3 type of update.

Will we see a return in our results anytime soon?

Thanks,

Justin

[RyanKent]

Your present approach is correct. Ensure all these pages are tagged as noindex for now. Remove the block from robots.txt and let Google and Bing crawl these pages.

I would suggest waiting until you are confident all the pages were removed from Google's index, then check Yahoo and Bing. If you decide that robots.txt is the best decision for your company, then you can replace the disallows after confirming your site is no longer affected by these pages.

I would also suggest that, going forward, you ensure any new pages on your site that you do not wish to index always include the appropriate meta tag. If this issue happens again then you will have a layer of protection in place.

[BrettKrasnove]

We're pretty confident thus far that we have flooded the index with about 15,000 low rank URLs all at once. This has happened once in the past a few years back but we didn't flood their index, they were newer pages at the time in which were low quality and could have been seen as spam since there was no real content but adsense so we removed them with a disallow in robots.

We are adding the meta no-index to all of these pages. You're saying we should remove the disallow in robots.txt so googlebot can crawl these pages and see the meta-noindex?

We are a very large site and we're crawled often. We're a PR7 site and MOZrank DA is 79/100. We have dropped from 82.

We're hoping these URLs will be removed quickly, I don't think there is a way of removing 15k links in GWMT without setting off flags also.

[RyanKent]

There is no easy answer for how long it will take.

If your theory about the ranking drop being caused by these pages being added is correct, then as these pages are removed from Google's index, your site should improve. The timeline depends on the size of your site, your site's DA, the PA and links for these particular pages, etc.

If it was my site I would mark the calendar for August 1st to review the issue. I would check all the pages which were mistakenly indexed to be certain they were removed. After, I would check the rankings.

[BrettKrasnove]

Hi Ryan,

Thanks for your response. Actually you are correct. We have found some of the pages that should be no follows still indexed. We are now going to use the noindex, follow meta tags on these pages because we can't afford to have theses pages indexed as they are particularly for clients/users only and are very low quality and have been flagged before.

Now, how long until we see our rank move back? Thats the real big question.

Thanks so much for your help.

Justin

[evolvingSEO]

That's a great answer Ryan... I wonder, just out of curiosity, if it wouldn't hurt to look at the cached version of the pages if they're indexed?  I'd be curious to know if the date they were cached is right near when the robots.txt was changed?  I know it wouldn't alter his course of action, but might add further confirmation that this caused the problem?

[RyanKent]

Justin,

Based on the information you provided it's not possible to determine if the robots.txt file was part of the issue. You need to investigate the matter further. Using Google enter a query in an attempt to find some of the previously blocked content. For example, let's assume your site is about SEO but you shared a blog article about your movie review of the latest Harry Potter movie. You may have used robots.txt to block that article because it is unrelated to your site's focus. Perform a search for "Harry Potter insite:mysite.com" replacing mysite.com with your main web address. If the search returns your article, then you know the content was indexed. Try this approach for several of your previously blocked areas of the website.

If you find this content in SERPs, then you need to have it removed. The best thing to do is add the "noindex, follow" tags to all these pages, then remove the block from your robots.txt file.

The problem is that with the block in place on your robots.txt file, Google cannot see the new meta tag and does not know to remove the content from it's index.

One last item to mention. Google does have a URL removal tool but that would not be appropriate in this instance. That tool is designed to remove a page which causes direct damage by being in the index. Trade secrets or other confidential information can be removed with this tool.