Why Is Google Tracking SERP clicks?

Garrett · #1 (**permalink**) 02-03-2004, 11:41 AM

In a recent phone interview Dan Thies informed me that Google has been tracking click throughs on their search engine result pages (serps) for some time.

According to Dan, "when you view the source code of a Google search result page, you can see the format of the links looks like this: <a href=http://www.example.com onmousedown="return clk(3,this)">. That "onmousedown" event calls a Javascript function which tracks the click. In this example, it would be a click on the #3 result."

This javascript creates a little image instance that tells Google that, say, link #3 was clicked. Dan thinks clickthrough tracking may become a part of the algorithm.

Quality Control. (Warning: speculation ahead.) It's possible that if your listing gets clicked more than the expected average for a given search term, it may get "some kind of mojo points." Google may measure this extra clicking as votes for your site and perhaps use this information in their algorithm.

"If nobody's clicking on the search results Google knows it's not delivering what they're looking for. What are people who are drilling down into page two or three clicking on? Google may float those up higher."

They can very easily identify the search results that fail to generate the click through rates that they should and begin adjusting their results that way.

Google used to have a feedback tab in their tool bar, and a link at the bottom of the results that asked "how are these results?" These methods of measuring the value of results required interaction from the searcher. Tracking clicks allows Google to do the work, and clicks tell a more complete story than direct user feedback.

As Dan said, "Real users don't complain - they just search somewhere else."

Why do you think Google's tracking clicks?

sparky_t · #2 (**permalink**) 02-04-2004, 05:25 AM

Hi all,

Although Google are obviously the GranDaddy of Search Engines, tracking clicks is indeed the perfect feedback.

Increasingly I have steered away from using Google, as I find more and more irrelevant results in the top spots. (admittedly I do still optimise for them, because of their status)

Google, like us all, still need to stay above water, and by providing the "most wanted" at top spot, will inevitably give the percentage of people the results that they ask for, rather than what the algorithyms are providing for them in the way of statistics.

If indeed it works they'll get my clicks, but only time will tell....

Dez · #3 (**permalink**) 02-04-2004, 07:34 AM

Quote:

Originally Posted by Garrett

In a recent phone interview Dan Thies informed me that Google has been tracking click throughs on their search engine result pages (serps) for some time.

According to Dan, "when you view the source code of a Google search result page, you can see the format of the links looks like this: <a href=http://www.example.com onmousedown="return clk(3,this)">. That "onmousedown" event calls a Javascript function which tracks the click. In this example, it would be a click on the #3 result."

This javascript creates a little image instance that tells Google that, say, link #3 was clicked. Dan thinks clickthrough tracking may become a part of the algorithm.

I haven't ever seen this in their click through URL's in the Search Results - are you sure that you're not talking about the AdWords advertising links?

If it's AdWords links, then of course they are tracking the clicks.

It may just be me, but I'm darned if I can see the clk(3,this) code you're talking about or anything similar?

Please provide more info?

Cheers,

Dez

rawdist · #4 (**permalink**) 02-05-2004, 04:07 PM

It's definitely there if you view the page source. I think this policy by Google makes perfect sense, after all if several people click on a result then chances are it's a good one. Hopefully it will stop things like price comparison sites coming out top for everything under the sun.

Electronic Perceptions · #5 (**permalink**) 02-05-2004, 05:12 PM

This actually makes a lot of sense to me because as I understand it, this is part of what determines AdWords positioning too, right?

On the downside of course, since displayed results won't have pay per click penalties, it could open up forms of abuse - or at least attempts at abuse - by owners trying to make their site look more popular with continuous clicking... or worse, setting up scripts to do it for them.

If they can prevent the fraud that's bound to arise from it then yes, I think it's a great idea.

- Kathy

chrisjaeger · #6 (**permalink**) 02-05-2004, 09:22 PM

I figured they were doing it all along, anyway. You just confirmed it.

It's not like rocket-science, right?

:)

Fruit & Veg · #7 (**permalink**) 02-06-2004, 11:14 AM

I'm with Dez, I don't see anything.

Can someone actually post a url of a search on Google so when I view source and do a Find for 'clk(3,this)' or similar - I'll actually see something?

minstrel · #8 (**permalink**) 02-06-2004, 11:30 AM

I haven't actually looked for it but what the previous posters are saying is not that it's visible on the page - it's in the HTML.

Click on View | Source in Internet Explorer or save the page to disk and open it in Notepad to look at the HTML code.

Fruit & Veg · #9 (**permalink**) 02-06-2004, 12:28 PM

Yeah, I know that. I just require visible proof and I ain't seeing it at the moment.

Garrett · #10 (**permalink**) 02-06-2004, 01:23 PM

I did this search:
http://www.google.com/search<br /> ...8&q=webpronews

Then viewed source on the results page. Then I did "ctrl f" and looked for "onmousedown."

You should find the proof that way. If you don't see it, it could be that Google's not tracking clicks universally.

Here's the incriminating code I found on the above search:

<a href=http://www.webpronews.com/
onmousedown="return clk(1,this)">

Is there anyone else who's not seeing the click tracking on the results pages?

Fruit & Veg · #11 (**permalink**) 02-06-2004, 02:05 PM

Interesting. This is what I originally see in Firebird;

Quote:

<a href="http://www.webpronews.com/">
WebProNews - Breaking eBusiness News</a>

And this is what I see IE6;

Quote:

<a href=http://www.webpronews.com/ onmousedown="return clk(1,this)">
WebProNews - Breaking eBusiness News</a>

Adwords are tracked via a 'num=7' being added to the end of the clickthu url, so it does look like Google is tracking clickthru activity in the natural SERP....just not in Firebird though.

bebarrett · #12 (**permalink**) 02-06-2004, 03:17 PM

I thought they did this all along.

On another note, reminds me of seeing a site that had all its internal links go through yahoo's SERP script. Of course, I had the idea to do it earlier, but my sites place well anyway, and I wouldn't want to jeopardize that by being caught.

smakyyy · #13 (**permalink**) 02-07-2004, 12:36 AM

yes- i see the on mouse down - but how does this play into the algorithm - how many clicks determie what is good- any thoughts?

Mel · #14 (**permalink**) 02-07-2004, 11:43 AM

If Google has a way of incorporating this data into the algo, it is the kind of feedback that search engines love,but it has to be taken with a grain of salt.

If for a particular search term the majority of searchers do not click on the top say five results then it is indicative of those top five results not really being what the searchers were looking for, which goes to the relevancy of the results.

But the decision to click on a result is based on more than just the position on the page but by how well written the page title and snippet are, and thus this sort of feed back could be highly influenced by these factors which have nothing to do with relevancy.

A nice feedback tool though.

OziBear · #15 (**permalink**) 02-09-2004, 07:42 PM

In the Aussie version of Google they have opted to use a tracking URL variable rather than the Javascript model as discussed above. Check out http://www.google.com.au and run a search for your keywords of choice.

The anchor tags for SERPs include
http://www.google.com.au/url?sa=U&st...age.htm&e=7249

I can't explain why Google has elected to run this style of tracking in Australia, but it does prove Google is serious about tracking CTs on SERPs.

One alternate view to those expressed above could be that Google is tracking CTs so they may lower rankings on popular sites (getting the clicks) so those sites are forced to take up AdWords to maintain traffic.

Conficio · #16 (**permalink**) 02-10-2004, 08:54 AM

Hi everybody,
I was wondering for a long time, why search engines do not track such things. However, it is not so strait forward as it might seem, to use this kind of data.

I'd argue that in the first place, the algorithm is self feeding. Off course people click on the first entries the most (that is why everybody around this water cooler wants to be in the top 5 / 10 /20). But beyond that the interesting thing is the entries on page 2, 3, 4, ... that still attract attention from the users.

In my opinion the real value lies in timing of events. One click doesn't tell me much about the relevancy of the page. It just says that the short description did not repel the user. But if the same user comes back within a few seconds, then I can conclude he didn't really find what he was searching for and moves on.

I'd say the fastest return (back to the google search page) would stem from any kind of misrepresentation or from an ambiguous search term such as apple (as in apple computer or apple the fruit), where the result page just falls into the wrong category. However it could also stem from someone trying to return to some website she remembers and does remember the general appearance (colors, images, etc.) and theme, rather than the name.

Another pattern that I would be interested in, are people that click on a few entries and then revise their search (add or exclude terms). This would indicate to me, that the clicks they made didn't resemble the results expected. This also leads to the clustering technologies used by some advanced search engines (dogpile, Kartoo, etc.), each cluster being a revised sub search.

One remark about scripts (sh, pearl, etc.). They should not influence the results, because scripts usually do not interpret the JavaScript, like a browser does. Has anyone seen a script engine that does interpret JavaScript? As the search engines are blind for JS generated menus etc. so are scripting engines for this code. Anyone proofing me wrong?

I'm saddened, that they obviously do only support this kind of function for IE. This should skew the results towards the standard consumer users (and exclude all Linux, BSD users). I guess the community using alternative browsers is more filled with nerds, researchers or professionals, because they show a more sophisticated taste in browsers and look for certain features. But I might be wrong.

It must be certainly fun to mine all this data. For me to make sense out of it, I'd have lots of focus groups (and research) to get comments on, why people click in the pattern they click. Unfortunately this kind of thing is very expensive. However it would be certainly better than just guessing.

I'm very weary about algorithms, that just measure popularity. Because this is grossly unfair to the new comer site as well as it does severely limit diversity. It leads to the effect that you wake up in your hotel and have breakfast and can't remember what city you are in, because they all look alike. This kind of world is an awfully dull world in my opinion.

Thanks for bringing up this very interesting topic

K<o>

OziBear · #17 (**permalink**) 02-10-2004, 10:34 PM

Conficio stated:

Quote:

One remark about scripts (sh, pearl, etc.). They should not influence the results, because scripts usually do not interpret the JavaScript, like a browser does. Has anyone seen a script engine that does interpret JavaScript? As the search engines are blind for JS generated menus etc. so are scripting engines for this code. Anyone proofing me wrong?

The JS contained in the Google pages actually invokes a script function when clicked which performs the following:

Code:

new Image()).src="/url?sa=T&start="+n+"&url="+escape(el.href);

On initial inspection this JS seems like it actually calls the server to serve an image. In fact an image is not called, but the server side script (/url) does run.

This script would most likely pickup the URL values and store them in a db or a log file. For example, the "start" variable is the (ranking) number which the link appeared in the SERPs. Google could compile stats on how many position #2s are clicked or #3s are clicked, etc. Google would also record the URL sent in the href element to generate stats like how often has this URL or domain been visited through Google SERP clicks.

This technique has been used for years in many instances, one which is most common being the tracking of bulk email messages. In HTML format messages (sent by spammers or legits) you will notice in the more sophisticated messages there is usually an IMG tag that doesn't actually call an image, but has something like SRC="php_OR_asp_OR_cgi" as what would normally be the URL of the image.

This allows bulk mailers to track how many % of a campaign has been opened (or re-opened), how many instances of the message were forwarded and opened by others (and who'd forwarded to them) and with some smart CSS this technique can be adapted to track % of messages that were printed (signalling these are really hot prospects, triggering follow-up messages or phone contacts). Those evil spammers especially like this technique, as they can verify with great certainty if a mail account is active, as someone is actually opening the messages. SIBEBAR: to avoid this type of tracking, download your mail, disconnect from ISP and then read messages.

I suspect Google is only compiling data at this stage, but with a considerable amount of data collected from weeks/months of users clicking, Google could start to draw conclusions from the stats that only we eMarketeers would dream about having access to. How they eventually (or if) Google ever applies this data to anything useful remains to be seen.