What's Cooking At MSN AdLabs?

By Aaron Goldman, VP Marketing & Strategic Partnerships
Appeared In MediaPost’s Search Insider

In its effort to entrench itself in the search community, MSN has made significant investments not only in technology and personnel, but in the area of research as well. One such example is Microsoft adCenter Labs.

AdLabs was created in early 2006 but has been steadily expanded and improved over the course of the year. Today it boasts some of the most innovative free -- emphasis on free -- research tools available to search marketers and is a welcome breath of fresh air.

Let's take a look at some of the tools available in adLabs and potential applications:

Keyword Group Detection. By extracting themes from a specific query, this tool provides terms that are relevant or similar to that query. For example, when I entered my last name, "Goldman," it turned back terms like "Friedman," Cohen," Levine," and "Schwartz." Sounds like the roll-call at my synagogue. Interesting to see how it made that connection, though.

Keyword Mutation Detection. Now this sounds like something you'd find in a lab. Igor, it's alive, it's alive! This tool shows common misspellings and other variations of a particular query--very helpful when building out a keyword list and ensuring you've covered all the possible ways someone might spell your brand name, product name, etc.

Search Funnels. With the notion of the funnel being all the rage these days, this handy tool shows the specific queries immediately before or after the keyword you enter. Options include drilling down within the funnel (e.g. "cars" > "used cars" > "autotrader") and adjusting the number of subsequent queries you want to view.

Search Result Clustering. This tool semantically clusters results for a query in a number of different ways ranging from questions and answers to displaying relationships among people. For example, a search for "Ashlee Simpson, Jessica Simpson" reveals their relationship as "sisters" first, following closely by people who have had a "nose job."

Forecasting Search Volume Seasonality. While it doesn't quite roll off the tongue like Google Trends, MSN's version is decidedly more robust if potentially less directional, due to its smaller share of overall searches. For one, it shows absolute search volume for each query, whereas Google just shows positioning relative to the axis. MSN also shows actual vs. forecast volume. To test the accuracy of the tool, I entered the keyword "tax" and, sure enough, it showed huge spikes from January through April.

Keyword Forecast. Perhaps the most immediately actionable tool in adLabs, the Keyword Forecast allows you to view actual impressions for a query by month for the past 12 months. It also overlays the age and gender distribution of the individuals who submitted that query. And, to top it all off, the results can be viewed in pictures or text for easy export into PowerPoint or Excel.

Detecting Online Commercial Intention. Now we're talking. This tool predicts the intent of each query based on the likelihood that it is commercial or noncommercial. It also ranks specific Web sites and further delineates commercial sites by informational or transactional. For example, a search for "sports" showed a high probability for noncommercial queries, while "cbssportsstore.com" revealed a high probability for commercial-informational. Now, while the Goldman Gut-Feel-o-Meter could probably return the same results for 9 out of every 10 queries using this tool, there's no question citing MSN adLabs at the bottom of a client reco deck is better than GGFOM.

A word of caution: it is not recommended to use adLabs data alone to make any radical changes to current campaigns or search strategy. The information in adLabs is based on a sampling of data--per MSN, currently 4.7 million keywords and growing--and should therefore be used directionally. While that number may seem like a lot, as you begin to play around with the tools, you'll notice that many long-tail queries and smaller Web site domains are not represented.

Nonetheless, as my dad once told me, "Some Data is better than no Data." Now, at the time, he was referring to a "Star Trek" episode that had only a brief appearance by everyone's favorite android, but the point remains, the information that can be culled from adLabs is a far cry from what was freely accessible to search marketers at this time last year.

If we're ever going to truly evolve the search platform or, to use a popular buzzword, cross the chasm (although I prefer ford the river--Oregon Trail, anyone?) from pure direct response to holistic brand marketing, we need tools like MSN adLabs to help us flesh out the insight into consumer behavior that is inherent in each query.

I applaud MSN for cheffing up free tools that will empower agencies and marketers to traverse the trail of insight. And I encourage the other top search providers to follow suit and up the research rations from meager to filling, so we can set a grueling pace of innovation and avoid data-dysentery along the way. Until then, for the most robust search info no money can buy, it's Redmond, ho!