What Was Your Intent With That Query? Search Engines Want To Know

A few months ago I wrote about the ways in which Google could measure traffic patterns and how they might be using those patterns in their search results. Many in the seo world would agree that Google, as well as the other engines, are using the information they collect to help determine what pages to display to a given search query. Recently I came across a few quotes from Google and MSN employees that would indicate our traffic patterns do indeed affect search results.

Before I get to the quotes I want to mention how I came across them. I’ve always believed that the more you know the better you can make decisions. I think it’s good to read from a variety of sources, and you never know where you’ll find something you can put to use in your business. My understanding of European geography, for example, came about studying history. You read enough about where Napoleon marched and after awhile you get to know the lay of the land. Understanding art and where the predominant schools were located during a given century helps you understand what countries were economically and politically strong during those same periods of time.

The information about search engines measuring traffic came about while reading PC World magazine. Not a far stretch from search engines and the web I admit, but when I started the article I certainly wasn’t expecting to find quotes from people working for Google. Microsoft, maybe, but not a director in the online services group responsible for Live Search. The article in question was one of those prediction articles on where technology might be headed and this one particlular part focused on search engines.

Search Engines Want To Understand Your Intentions

Back to the quotes. I know that’s what you want to see. First from Peter Norvig, director of research for Google talking about how Google returns results to search query.

“We want to do a better job of understanding the user’s intent and the content provider’s intentions,”

and

“We mostly rely on matching keywords, but we’d like to get closer to matching the intent.”

Another is from Adam Sohn, the aforementioned Microsoft director

“If someone is searching for ‘Jaguar, the smarts to distinguish between ‘he’s looking for a car’ and ‘a big cat in the jungle’–that’s coming.”

Those quotes are all about determining the intent of the person typing the query into a search engine. You’re never going to determine intent from spidering and indexing web pages. If you want to know what my intention was when typing ‘bass’ into a search engine and whether I was interested in ‘bass guitars’ or ‘bass fishing’ you need to be tracking something about me. That’s where measuring traffic patterns comes in.

Some of this can simply be about personal search and I think in part it is. If I happened to be searching recently for electric guitars or Charlie Mingus a search engine could reasonably surmise I want information about the musical instrument and not the fish. Of course I would need to be logged in when searching for them to deliver personal search to me.

Beyond the above searches they might tie my past searches into the searches of others and determine that maybe 73% of people who had searched the phrase ‘firewood delivery’ also spent time on sites for bass guitars and use that to deliver search results for the query. Traffic patterns while more geared toward helping with relevancy could be used to help understand intent. Just show a mix of results for the query ‘bass’ and see what gets clicked more. If bass guitars are clicked 6 times out of 10, maybe the first page of results should be 6 guitar pages and 4 fishing pages.

Understanding Intent Through Social Communities

Another way in which search engines are moving away from mathematical algorithms comes from the following quote by Mr. Sohn

“Over time, especially with video, there will be this social input, where people add tags to other people’s video. Then you get this sort of community-reinforced set of searchable attributes.”

Tagging is much like adding meta tags to a page except that the tags are added by those who didn’t create the page. Still open to manipulation, but more difficult than manipulating meta tags. Social networks are a way for search engines to learn from us more about what we want. Is it any wonder search engines are gobbling up social community sites left and right. Maybe it isn’t really about the advertising, but what the engines can learn about what really want to see when we type a query. It’s easy to determine a lot about a person by what they bookmark or tag and that information is available even to the public often enough. When you have the keys to the back door you can determine even more.

As I mentioned in the other post on traffic patterns, regardless of whether or not user behavior is currently built into the algorithms, the search engines are clearly collecting the information and there’s no reason to think they won’t be using it at some point. They probably are to some extent right now and there’s no reason to think they won’t use more as they collect more data about us.

Implications For Webmasters

What I think this means to webmaster and site owners is something we already know. The use of data outside of what’s spidered and index corroborates the need to build web pages and websites for people and not search robots. If the results are ultimately influenced by how people search and what their likely intent is, then it’s a good idea to please those people with our sites. If you want someone to actively tag your site or write a positive review you probably want to create a page useful for that person.

Instead of stuffing keyword after keyword into your page title, might it not be better to write that title in a way that makes someone want to click on the link. Google already uses click through rates in determining the cost per click of your AdWords ad. More clicks leads to lower cost per click and more times your ad is displayed. Is there any reason to think they won’t at some time, if they aren’t already, re-order organic results based on how often results get clicked and how long a person stays on the page after clicking.

Search is evolving to include more than algorithmic results. The algorithm will likely always be present, but there are now and will continue to be more layers on top of that algorithm. Google and Microsoft want to know what your intention was when you typed in that search phrase. I’m sure Yahoo and Ask want to know as well. The only way they can determine your intent when you search is by measuring what you do and where you, or someone they think is like you, go after they return results to your query. Google and Microsoft are saying as much when they say they want to understand your intent when you search. They are saying it when they say they want to present results based on your intent as well.

Download a free sample from my book, Design Fundamentals.

5 comments

  1. You can also note that you can ponder about the searcher’s intent when looking at your website stats to create content.

    Generally, the art of selecting the right keywords is built around the customers. Some keyphrases are obvious target ones, while some seem right, while being used by a slightly different audience. Different enough not to use your product or service.

  2. This reminds me of the Futurama episode where Professor Farnsworth said “fry” and the computer ordered him french fries and opened his calendar to friday.

    So far search engines have relied heavily on searchers learning how to search and find what they want, Which isn’t a bad thing, Though obviously they will want to improve thier search engine and part of this is going to be the ability to second guess a searchers intention.
    I just hope this doesn’t mean we will all have to learn how to search again lol.

  3. Doh, I don’t think I’ve seen that one :)

    Well, the search engines won’t make it harder to search (I hope). What they may, instead, is improve their interface to excite more precise queries (as per this interview with a search query researcher, Amanda Spink).

    Then again, there is Ask, who offered another method of searching, which might be helpful. I heard Google tried to implement the same algorithm (answering to questions) too, but I haven’t seen or heard about the results.

  4. I didn’t mention it in this post, but in the article both Peter Norvig and Adam Sohn agreed that improving the interface was something that would definitely be happening in the near future (12 – 18 months time). I don’t have the article in front of me now, but they want to be able to offer better recommendations and get all the different types of search inline. I think the idea would be to make it easier for you to search images, and the web, and video, etc all from a sngle interface. Something like that anyway.

    I’m still not using Ask regularly, but they do have some nice features. I wish they would let you reorganize their search tools between pages, thoug. There are some I like on the second page and some I’ll never use on the first page. Unless I’m missing it I don’t think you can move tools between the two sets.

Leave a Reply

Your email address will not be published.

css.php