6 replies to this topic

[Robmuller]

Hi

Since a few months I use Google Analytics as visitors tracker for my website. To use Google Analytics you have to install a script on every page of your site, so Google can track the behaviour of site visitors. Currently Google sends about 2,000 visitors per day and that number is increasing by about 25% each month. The refusal percentage (I do not know the English fraze, since I am using a Dutch version of Google Analytics) is 11%. This means that 89% of the visitors that Google send find that the site meet their Google searches.

My question: does Google use Google Analytics data in their algorithm? I would find this very logical since GA gives very detailed information about the response of visitors regarding keyword frazes they use to search with Google and the level of satisfaction (visitor seaches via Google, e.g. "barcelona restaurants" --> my site ranks high in Google --> visitor visits my site --> 89% stay for more than one minute and visit more than 3 pages --> Google measures this --> Google gives a higher ranking for the specific search the visitor keyed in.

Does this make sence? If it does, webmasters who really have a good site should install Google Analytics, so Google can take the visitors behaviour into their algorithm.

Best wishes

Rob

[MaKa]

This has been discussed in other threads on the forum. I had a quick look, but couldn't find the one I'm thinking of.

If I remember correctly the argument against Google doing this is that it's really easy to fake traffic and hence influence their positions. It would therefore not be a likely ranking factor.

Groetjes,

Marijn

- is it just me or are there a lot of Dutch people on the forums all of a sudden?

[torka]

If I remember correctly the argument against Google doing this is that it's really easy to fake traffic and hence influence their positions. It would therefore not be a likely ranking factor.

Not only that, but only a portion of the webmasters have GoAn installed, probably not even the majority. Google is unlikely to take into account (as a positive or a negative ranking factor) anything they can only measure for a minority of pages in the index.

--Torka

[Jill]

Since Google wants to be the best search engine in the world, I would imagine they would use all data at their disposal if they felt it would help to provide more relevant results.

[piskie]

Right on the nail Jill.

[nethy]

They also have a right to use the data any way they want, don't they? I mean isn't that the deal? paid tool for free in exchange for data?

I would imagine they are using the data to gain generic insights though about the relationship between a page, a search term and user behaviour. Maybe also specific.

[Randy]

It's one of those things they've never actually explained at all, let alone gone into any detail about.

However I would tend to agree with the above sentiments. To what degree they might use the information, no one knows. I would however assume they're using it to some degree. If not on a grand scale, at least to some degree.

As a quick for instance, let's say a site was using GoAn and the stats showed that 80% of those arriving by a specific keyword phrase converted to whatever goal the webmaster had set up. This could be a good indication to Google that this specific web site really is what people are looking for when they search for that phrase, so if it's not already ranking well for that phrase I can see how and why Google might give it a bit of a bump so that it gets better exposure.

By the same token, if they see from the GoAn status that 95% of those arriving at the same site immediately bounce from the site after just a few seconds, it might be an indication that the search algo has made a mistake in ranking the site so well for this search phrase.

I've not heard anyone putting forth such a scenario, however it wouldn't surprise me if it starts happening somewhere down the road. If it's not happening already.

My personal hope is that Google continues to get more involved in all things marketing related, possibly releasing some general info about how web sites are performing on the conversion side of things. They should end up having a large cross section of data at their disposal and it would be nice to have some real benchmarks for people to measure their own marketing/conversion success against.