Why doesn't my click in a search result go directly to the destination?

Before I start, I want to point out that while I'll use Google (and as you'll see, Starbucks) in my examples below, what I'll be discussing applies pretty much to any search engine and any search result.

In the abstract

I'm going to side-step the specific site issue (Google UK versus Google U.S.) primarily because I don't know why they would be different. However, I do know that Google is constantly changing things and they've been known to roll out changes at different times to different locales.

It might all be different depending on whether you're logged into your Google account, which browser you're using and a raft of other things that Google pays attention to as it tries to get you the best results for your search query.

And, ultimately, that's what it's all about: Google trying to get you the best search results that it can.

Transparent redirection

Let's start with a simple search in Google using Internet Explorer 9 for the word "coffee". My number one result is, naturally, Starbucks:

And that looks like a link to http://www.starbucks.com. In fact, the tooltip that appears when I hover the mouse pointer over the search result would seem to indicate the same:

And yet ... if you right-click on the search result and select "Copy shortcut", the result is decidedly more complex:

http://www.google.com/url? ... &url=http%3A%2F%2Fwww.starbucks.com%2F& ...

I've replaced a bunch of stuff in that URL that is simply Google's own tracking information. The key points are these:

• When you click the search result, you don't go directly to the search result at all, but to "http://www.google.com/url"

• One of the parameters passed to that page is url=http%3A%2F%2Fwww.starbucks.com%2F, which is an encoded form of url=http://www.starbucks.com/, the place that you really want to go.

In my experience, going to that long google.com URL produces what's called a "302 redirection" to "http://www.starbucks.com". That's a function of the HTTP protocol which tells your browser that the page that it's looking for is actually at this other URL. That "http://www.google.com/url" didn't generate a page to be displayed at all.

In this case, no additional pages are entered into your history and the browser's Back button will return you to the search results.

Not-so-transparent redirection

In one of your cases, because a browser "Back" operation doesn't take you back to the search results, that means that the sequence is actually slightly different.

• When you click on a result, you're taken to http://www.google.com/url as well, but this time for some reason, an actual page of HTML is created and fed to your browser. No HTTP protocol-based redirection occurs.

• That resulting page then includes information that in essence says, "Take me to http://www.starbucks.com immediately."

• The browser then loads the destination, starbucks.com in my example, as a replacement page.

The net result is the same - you end up on the destination page that you clicked in the search results.

The mechanism of how you got there is different, and in this case, it did cause an "interstitial" page to be entered in to your browsing history. Hit Back and that's the page that you're taken to. Because that page says, "Take me immediately to starbucks.com," that's what happens - you click Back, the interstitial appears briefly, if at all, and you land back on Starbucks.com as if you'd never hit Back at all.

Why and when Google might use one technique over the other is unclear, but obviously, they do.

Why redirect at all?

So why would a search engine redirect instead of just letting you click directly on the link that you want to go to?

It's actually pretty simple, really.

They want to know what link you clicked on.

Redirection is Google's way of being notified that you clicked on that specific link and not one of the others.

Why do they care?

Because that tells them that you determined that that specific link, more than any of the others, was a possible match for what you were searching for. In my case, when I search for "coffee" and then click on the Starbucks link, that tells Google that I feel that Starbucks is somehow related to coffee.

Now, you might be saying, "Of course, Google knows that Starbucks is all about coffee".

Yes and no.

Starbucks' own website is pretty clearly about coffee and Google can make a lot of inference about the relationship right there, but not all search results are that obvious. Particularly when the relationships aren't that clear, understanding what people actually click on is one way that Google develops better search results.

Note also that I emphasized that Starbucks was my number one search result.

It might not be for you.

Google uses many pieces of data to personalize the search results, including location, browsing history, which search results you click on when you search for something, and much more...

I visit starbucks.com often enough that Google knows that it's a pretty important site to me. But if you've never visited that site, it might not make that inference. I also live near Seattle, where Starbucks headquarters are located. It's very possible that the same search performed elsewhere - perhaps in a location where (gasp!) there are no Starbucks - would produce very different results.

And, if you often search for 'coffee' and typically click some link other than the Starbucks link - perhaps the Wikipedia link - then over time, Google may well determine that when it comes to coffee, Wikipedia is more relevant to you than Starbucks.

It's all about better results

And that's really what it all boils down to. Redirection of search results clicks is Google's way of collecting information that they then use to improve the results of future searches.

And that improvement can be for all users of Google, or simply to make your search results better.