what does “(not provided) keywords” really mean

you might find this article useful, i talk about this issue and how people can tackle it.

http://atladasmedia.wordpress.com/

The amount of people logged in will no doubt vary from site to site, with a majority of visitors to your own being particularly web savvy, probably each having a Google+ account and remaining logged in.

I see this becoming a trend, sites that have a particularly technical demographic (computer retailers, forums, tech blogs) experiencing a higher percentage of logged in users than, say, mumsnet (possibly a bad example).

I also have the feeling, because of the nature of the technically savvy web users, it may cut out more of the medium and long tail rather than the high volume glamour terms, but this remains to be seen.

We have been in the 1-3% range of daily traffic coming in on the (not provided) keywords since they announced it.

Yesterday though this doubled and we moved up in the 7% range coming from the (not provided) keywords.

Not sure why the big jump yesterday. Curious to see if it stays in that range now.

I said all along that the claim that only a “single figure” % of people search logged in.

Especially when you consider anyone who is logged into

Gmail
Blogger
G+
Picasa
Youtube
Android Marketplace
etc

is auto logged in on search… fairly large chunk of the web and a chunk that is only going to grow!

Hi Dave,

How did you determine those figures? http sniffing “hack” or something else?

we had to create an hack

Are you going to write about it?

David, these results were only from people who used Google search engine to get to your site, right?
I’m having a hard time reading the results and accepting them to be true. As you already said, it was a hack by your team. And since i can’t duplicate that, i cannot really believe the results. As much as i would like to of course.

@beany at present Google is only reporting (not provided) what I’m doing is detecting whether a user is logged into Google when they come to my site

Yeah, thanks that much i did understand.

So what you’re saying is that out of 8.800+ visitors from Google, 25% was logged in, right?
Something else you’ve said in the article is that when you’re logged in, search for something on Google, the SERP is not SSL for you. Right? Because i got the same thing. Does this mean the implementation has not been worldwide yet, or just some other criteria/setting?

crap times, let’s hope Google+ falls on it’s arse now otherwise we’ll have no data

@Dave – Do you have an example of the SSL only Google URL yet? Been searching for one but no success just yet

I am not MR SEO KING, but why not use another Analytic tool? When I was doing SEO full time back in the days like 15 years ago we had non google tools to know our traffic. I think we need to get off wall street and then go occupy Bing.com or Search.com or some other search engine, and I think with these other movements right now is the perfect time to move people from google to some other search engine. Let the people know how much google knows about them, google everywhere, your phone, gps, text, email, search, video, social media. They are way too much involved in our lives. They are even telling us what we want to see based on your patterns and claiming it to be a benefit to us. I do not search google anymore myself. If apple has a search engine I would rather use it and I am a PC user. I know they have I think safari and I have it on my laptop.

That’s the plan, even thinking of going old school to get the keyword data but the simple fact that Google Analytics doesn’t have any strong competition on a price comparison means it’ll still be used, even by me for the moment.

Also welcome to the 21st Century where brands kinda force their views upon us (and the Big G believes brands are the light within the ‘cess pool’ that is the web)

@beany Does this mean the implementation has not been worldwide yet, or just some other criteria/setting? I think it’s not rolled out worldwide and it’s only on the .com, we are testing again on another website totally on seo related will update wheni have data

@dudibob you get a normal Google search referrer but the q=&esrc=s

Google “Do No Evil” is an empty slogan

David hi, I was just thinking the following. Let us say that your 100% visits looks like this:

33% Other search engine, sources
33% Google search
33% Social media

We will still now the 66% of the keyword queries so maybe we can modify the “unknown” 33% on the basis of SM and other sources?

Just food for thoughts

7.18% of Total Visits on a client that gets over 22,000 Organic searches per month
in the off season and about 30,000 in the peak season.

I’m getting 10%. It’s a bummer.

The 25 pc sounds low to me considering the sites mentioned above will log people on to search auto.
I’m not really understanding the implication though. That’s what I was hoping to see when I clicked on the link.

I have to ask one simple question, and I mean it is simple. When will the SEO people, the uses of google, etc all jump ship from them to use search.com, bing.com, yahoo.com, etc. I know as a user I do not want the machine to tell me what I want to see. I really do not. I am not an SEO guy, but a user. I want to find things naturally. I do not want it to build a pattern by me. I guess I have more than one question. Why does everybody use google analytic? You are just feeding google with data.

Hi James,

I’d hazard a guess that it’s because Google seems to be most efficient in returning accurate, useful search results, is non-cluttered and Analytics is free and engaged in constant development.

Economically, ignoring Google and concentrating on the search engines where the vast majority of USERS aren’t makes no sense.

Just my 2 pence.

Matt

Some guys wanted to know how the hack works. Let me quickly explain it and also highlight an issue with it:

Dave calls an image in the source code of the page that can only be accessed when you’re logged into Google Mail:

[img height=”1″ width=”1″ style=”border-style:none;” alt=”” src=”https://mail.google.com/mail/photos/img/photos/public/AIbEiAIAAABECIfh2736y4XbtAEiC3ZjYXJkX3Bob3RvKig2OTAzMDA3NDAwNmUyZDI3MDNkYjAzYzhjMjEzZWUxZjIyZjNiMjNiMAGGFCA0egzvk9Ic5gy6pk6QEOA3sA” onload=”gli(true)” onerror=”gli(false)”/]

If the image loads fine it will call the function “gli(true)” (onload). If the image cannot be loaded it will call the function “gli(false)” (onerror).

The JavaScript function ‘gli’ will set a custom variable and log a custom event in Google Analytics:

function gli(val) {
_gaq.push([‘_setCustomVar’, 3, ‘Logged in to Google’, val ? ‘Yes’ : ‘No’, 2]);
_gaq.push([‘_trackEvent’, ‘Custom’, ‘Logged in to Google’, undefined, val ? 1 : 0, true]);
}

So depending on the image the custom variable ‘Logged in to Google’ will either be set to ‘Yes’ or ‘No’. That’s it. Pretty simple at the end but a nice hack.

There is an issue when you have a Google account but doesn’t use Google Mail. In this case the image won’t show up either and the gli-function puts me in the ‘No’ (not logged in) bucket.

What does that mean for the stats above? The number of logged in visits could even be higher.

But I’m not a JavaScript pro. So feel free to correct when I’m completely wrong.

very useful article. i really like it.

I have to agree that google search is deteriorating for me and I often look at other engines too. Seo takes too much precedence spend more time making your sites useful and better and the seo will often take care of itself.

Yes, stay focus on doing the right and making it logical and it will all flow nicely, but it is time for a social movement away from google. Nobody, no GOVT, no business should know my every move. I feel like a rat now. At least feed me some cheese.

In fact the problem is not GA, it is that the SSL does not provide the referrer as far as I understood, so other analytical apps would not work also

Agree strongly. Too many businesses rely on google. I think I’m going to start using other analytical tools. I had a site again different site just drop from planet google. A good site not really high but some useful info , google algorithm decides all of a sudden not to rank it at all ! And
@matt I don’t think the results for short tail results are very good from google ! Anyone else get better results for a term ? I’m going to do some real tests

Sites I control started out with a low percentage of logged in users. However, it is increasing each week. I have not seen anything over 20% yet like you have, but some large traffic sites are nearing the 20% mark.

Notice about this feature is spreading and since “privacy” is attached to it as a benefit more and more people will search while logged in. Who would not if doing so protects their privacy? The masses scream at Facebook for poor privacy settings, yet post pictures of them puking in the front yard at the last party they attended and have a public profile. The masses have an odd relationship with the Internet. They love being on it and sharing information yet want their privacy protected.

(not provided) has become a good search term for you David as that is exactly what I just searched for to see what it meant, I thought my site was coming up with the keyword ‘(not provided)’, as yours does and I along with many other people have clicked the link how would that come up in Google Analytics?

Good point ! I guess it would show as not provided !

what does “(not provided) keywords” really mean

30 Comments

Get in Touch