Tag: algorithm

Google’s Adsense Algorithm

Posted by Robert Gentel – August 8, 2008

Google’s search algorithms get a lot of play and not enough people are paying attention to the fact that Google’s contextual ad network still enjoys a technical superiority to its peers that it long lost in the search relevancy algorithm. Simply put, some search competitors are doing a decent job with search relevancy but still seem to be nowhere when it comes to serving relevant ads.

I’d like to share some of my thoughts on the Adsense algorithm, which I will revisit in detail in the future. Given the secrecy of the sauce I will not try to prove what is and what is not the Google Adsense algorithm and will take the approach that any SEO worth his salt should and speculate as to what the algorithm should be, and what current technology is capable of.

At its simplest level I believe the algorithm should, and likely does, work like this:

  1. Attempt to determine page context and serve a contextually relevant ad
  2. Use clickstream data to determine what the user might be interested in and serve an ad that may not be contextually relevant.
  3. Use basic demographic data (e.g. geolocation) to attempt to target ad relevance to the user.

The premise is simple, the context of the page is a strong indication about what the user will click on and is the first priority of the algorithm. You may know that the user was interested in other, potentially more profitable, subjects but that the user is on that page now is a fairly good indication of what the user is interested in at that particular moment.

But then again it isn’t always the case, and clickstream data can help identify what the user is really interested in. For example, the user’s previous searches can indicate what is really meant for the query “apple”, but even more immediately relevant is that Google often knows where you were right before you got to the page. And with increasing frequency, it was Google itself.

This is the single biggest reason that clickstream data must be a part of the Google algorithm. It’s much easier to determine context from a user-input query. That’s why other search engines are starting to compete with Google in relevance on most queries. If Google knows what the user searched for before clicking on this page they have a variable that rivals page context in relevance to the user. If they know the user searched for “buy a green widget in San Diego” and landed on a general page about green widgets they would be foolish not to use the additional context they know about the user (the location specific subset that they are looking for) in their attempt to serve an ad the user is most likely to click.

The “session context” as I, as of a moment ago, like to call it in the clickstream would be weighed heavily with page context in my algo, and historic clickstream data would follow at a distance. If you know they area always looking for certain widgets and you don’t have a great ad for the page or session context then an ad about what the user has expressed past interest in is the next best thing. Google has a lot of clickstream data from their own web properties as well as others through sites running Adsense itself as well as their free log analytics service they provide to webmasters in exchange for the data. For example, they could know what you searched for on Yahoo when you land on a page with their own ads or log tracking and it’s precisely such examples that they can use to their benefit. Search history presents the easiest contextualization opportunities because the user has given the context in a string. Other clickstream data requires a lot more guesswork and for these reasons I think that Google should, and does, focus on mainly search related clickstream data. Given my read on their corporate culture, I’m not sure if they are doing this outside of their own web properties, as in my Yahoo example, but they should and I can’t imagine that they don’t for their own search engine.

Lastly you can throw anything else you know about the user. You have the IP and can map to geodata in a simple example, like showing the user an ad for a local restaurant. And you can even get fancy and use aggregate trends (e.g. people searching in a certain area might be in town for a certain reason, come up with your own specifics) and other logical deductions (i.e. “wild guesses” like searching in English from Mexico might mean you are interested in a hotel). I think focusing your efforts is a big part of the end result of this kind of work and believe that if Google uses any of this fall back data they do it simply. Why spend time on a complicated algorithm to generate poor guesses when you can spend more time nailing the real priorities like page context?

In another post, I’ll break down the on-page context algorithm possibilities but I’m out of time for today.

Google gives details on its customized search results

Posted by Robert Gentel – July 30, 2008

Google has been collecting searchers browsing and searching habits for years. They have their search logs, clickstream information for every site serving their AdSense ads, and for every site using their free web analytics program they have information on the browsing history for their toolbar users who do not opt out, and of all the users of their Web Accelerator proxy. And unlike many other companies that collect user data so Google actually uses their data in fundamental ways. So it comes as no surprise that they’d want to find ways to employ user information in their search algorithms. Clickstream data and folksonomy are some of the big areas that search algorithms are expected to use. Right now all major search engines use ranking algorithms that are primarily based on the Pagerank concept Google introduced and became famous for. They all use links on the web to establish authority, and no fundamental change has taken place in the evolving search algorithms in many years. They get better at filtering malicious manipulation and at tweaks that eek out relevancy but nothing groundbreaking.

So authority based on clickstream analysis and social indexing seemed like good ways to use data to further diversify the effort to allocate authority to web pages. What Google learned early is that they needed scale, and their initial data efforts (things like site ratings by their toolbar users) didn’t end up in their search algorithm. Folksonomy and social indexing doesn’t yet have enough scale to rely on and has potential for abuse, but the clickstream has scale and is harder to game given that traffic is essentially the authority and the people gaming the authority want traffic. So if they need traffic to rank well to get traffic then there’s a significant challenge to those manipulating rankings because they need the end for their means.

But Google is cautious with their core search product and has tweaked their algorithm very conservatively. And it has been hard to tell just how much clickstream data was playing a role in their search results and will continue to be as long as it’s such a minor supplement to their algorithm. Today, Google has posted a bit of information about this on their official blog in their efforts to shine more light on how they use private data. You can read about it here in full but the basics are no big surprises:

  • Location - Using geolocation or your input settings they customize results to your location slightly. My estimate is that they are mainly targeting searches with more local relevance. And example of such a search would be “pizza”. “Pizza” is more local than “hosting” and can benefit greatly from localization. Hosting, not so much.
  • Recent Searches – Because many users learn to refine their searches when they don’t find what they are looking for this session state is very relevant data. A user who’s been searching for “laptops” for the last few minutes is probably not looking for fruit when they type in “apple” next. They reveal that they store this information client side and that it is gone when you close your browser but because they mean cookies anyone who’s been seriously looking under the hood already know this.
  • Web History – If you have allowed them to track your web history through a Google account they use this to personalize your results. They don’t say much of anything about what they are really doing but this is where a the most can be done and there are far too many good ideas to list here. Some examples would be knowing what sites you prefer. Do you always click the Wikipedia link when it’s near the top of the search results even if there are higher ranked pages? Then they know you like Wikipedia and may promote its pages in your personalized results. Do you always search for home decor? Then maybe they’ll take that into consideration when you search for “design” and not give you so many results about design on the web. There are a lot of ways they can use this data, and this is probably an area they will explore further.

In summary, right now I’d say their are mainly going with simple personalizations and not really employing aggregate data and aggregate personalization to give aggressive differences. They are careful with their brand and will use user history with caution. After all, if your the use of your results lead to less relevance they fail and because personalization can be unpredictable (there must be some seriously weird browsing histories out there) they are going to be cautious and slight with this.