Is Google able to determine duplicate content every day/ month?

[Martijn_Scheijbeler]

A while ago I talked to somebody who used to work for MSN a couple of years ago within their engineering department. We talked about a recent dip we had with one of our sites.We argued this could be caused by the large amount of duplicate content we have on this particular website (+80% of our site).

Then he said, quoted: "Google seems only to be able to determine every couple of months instead of every day if the content is actually duplicate content". I clearly don't doubt that duplicate content is a ranking factor. But I would like to know you guys opinions about Google being only able to determine this every couple of X months instead of everyday.

Have you seen or heard something similar?

[Dr-Pete]

Sorting out Google's timelines is tricky these days, because they aren't the same for every process and every site. In the early days, the "Google dance" happened about once a month, and that was the whole mess (index, algo updates, etc.). Over time, index updates have gotten a lot faster, and ranking and indexation are more real-time (especially since the "Caffeine" update), but that varies wildly across sites and pages.

I think you also have to separate a couple of impacts of duplicate content. When it comes to filtering - Google excluding a piece of duplicate content from rankings (but not necessarily penalizing the site), I don't see any evidence that this takes a couple of months. It can Google days or weeks to re-cache any given page, and to detect a duplicate they would have to re-cache both copies, so that may take a month in some cases, realistically. I strongly suspect, though, that the filter itself happens in real-time. There's no good way to store a filter for every scenario, and some filters are query-specific. Computationally, some filters almost have to happen on the fly.

On the other hand, you have updates like Panda, where duplicate content can cause something close to a penalty. Panda data was originally updated outside of the main algorithm, to the best of our knowledge, and probably about once/month. Over the more than a year since Panda 1.0 rolled out, though, it seems that this timeline accelerated. I don't think it's real-time, but it may be closer to 2 weeks (that's speculation, I admit).

So, the short answer is "It's complicated" I don't have any evidence to suggest that filtering duplicates takes Google months (and, actually, have anecdotal evidence that it can happen much faster). It is possible that it could take weeks or months to see the impact of duplicates on some sites and in some situations, though.

[Martijn_Scheijbeler]

Hi Donnie,

Thanks for your reply, but I was already aware of the fact that Google had/ has a sandbox. I had to mention this within my question. I'm looking more for an answer around the fact if Google is able to determine on what basis if pages are duplicate.

Because I saw dozens of cases where our content was indexed and we linked/ linked not back to the 'original' source.

Also want to make clear that in all of these cases the duplicate content was in agreement with the original sources just to be sure.

[SEODinosaur]

In the past google had a sandbox period before any page (content) would rank. However, now everything is instant. (just learned this today @seomoz)

If you release something, Google will index it as fast as possible. If that info gets duplicated Google will only count the first one indexed. Everyone else loses brownie points unless they trackback/link back to the main article (first indexed).