Judging from the emails, Tweets and requests for more coverage on how and why human judgment can and must play a role in our search results, I can say my last post on social search created quite a buzz in the blogosphere. I’m happy that an invigorating exchange has followed this post and even more pleased that it has us all  thinking about what I view as the perfect fit between our mobile devices (personal) and our increasing demand for genuinely useful (personalized) results on the move.

I have had this trend high on my radar for several years, a passion that received its first outlet and accolades when EContent magazine gave me free reign to write an in-depth cover story on the state of Social Search (which appeared in the November 2007 issue). I loved researching and writing the article, a work I still regard as one of the best in my career. The good news: I’m told the article had – and continues to have – an impact on the content industry. The not-so-good news: When I wrote the article I was disappointed that so few companies “got” mobile, and today – nearly two years later – only a handful of companies have really have caught on.

To be fair, I wrote the article, aptly titled Teams Work: Social Search Gets Results, before companies such as abphone, ChaCha, Hiogi and Taptu broke on to the scene with strategies that draw on social search approaches and algorithms to improve mobile search results ranking and relevancy. I invite you to read my comments in new-release white papers from abphone and Taptu, and MSG’s own soon-to-be released assessment of the user experience delivered by search engines (among them ChaCha).

But progress is progress, and I am confident that more online social search companies will sharpen their focus on mobile as the advance of mobile social networks and other communities pushes people-powered mobile search to the top of the agenda this year. (In fact, recent reports/stats on Twitter, Facebook, and my own discussions with social networking companies confirm an exciting new trend: We have begun to search in communities – and today the number of queries even exceeds our searches in Google. Connect the dots and a game-changing search paradigm emerges. The power of people + the power of mobile = a power shift in favor of new mobile players who harness the wisdom of mobile crowds.

Where does all this leave mobile operators?

Well, just a week ago I would have said they can watch if they are stupid, and open up their analytics to third-parties (to give critical context to search results) if they are smart. Now, I can say they have a third and far more lucrative option: They can offer a people-powered search service provided by an infrastructure vendor (building on existing front-end analytics capabilities that capture the interests, passions, and profiles of users) to deliver subscribers genuinely useful and relevant search results. And if mobile operators chose to bolt on the mobile advertising module, they can position themselves to offer paid search advertising from a variety of ad networks and keep the lion’s share of the revenues.

You’re reading it in MSG first: Watch for a mobile infrastructure provider to launch a solution that joins existing analytics at the front-end with search and advertising, potentially allowing operators to be equivalent to Google. And watch a mobile operator customer to implement it this year.

I am under NDA and therefore cannot disclose the name of the vendor company, but I can share the details of the search technology and the recent results of a trial test the vendor company conducted with a mobile carrier (a pilot that effectively proves the business case for mobile search powered by mobile searchers (subscribers) and enhanced by operator analytics).

Let’s examine the business case for operator-managed mobile search and advertising.

First, mobile search is not rocket science, nor is it a service that only search engine providers can offer. Search can be broken down into two basic tasks: Indexing the Web to include domains and fresh content in search results; and developing algorithms to assist in ranking and rating the results. It’s not impossible and white-label search providers have shown how this can be accomplished. (They just lost the plot when it came to finding and indexing the content on the wider Web, which is why white-label has lost momentum.) However, for reasons I outline below, mobile operators are positioned to go one better than the solutions that have gone before.

Second, PageRank is not the only game in town. There are many techniques available, not just PageRank, a one-size-fits-all approach pioneered by Google (that as I point out in my post suffers from serious limitations in the mobile space). Operators, because they are close to the customer, can harness social search concepts to improve the experience and the search results. (One to consider is BrowseRank, a new algorithm developed by the vendor company that represents a vast improvement on PageRank because it focuses on the pages people click most often, and correctly assumes that we click these pages because they offer the content and answers we value most.

Third, mobile operators alone have access to the data that matters - directly from us. They know our location and context; our profiles and purchases; the sites we browse; and the search results we think rock (!). As a result, mobile operators don’t have to index the whole of the Web because they know from their usage logs the sites we browse and click (not as individuals, but as a group), and can simply make sure these domains figure prominently in the mobile search results.

Finally, we don’t need the whole of the Internet on our phones. If people really do want/demand the whole Web on their phones, then it makes sense to leave mobile search to the search giants who can handle the mammoth task of organizing the world’s information. But do we want it ALL? Not if we consider a recent trial involving the vendor company and a North American carrier. As the source put it: A review of a month of mobile search usage and queries revealed that “millions of users had only visited 34,000 domains.” Drilling down, the source found that a whopping “96 percent of transactions were covered by 14,000 domains.”

Hmmm… So, is there really a Long Tail in mobile search? Or have bad experiences – poor content adaptation, broken links, and other shortcomings – already dampened our enthusiasm for the mobile Web? Or are we just gravitating to our favorite destinations (such as social networks) on the move, and leaving the Long Tail searches to our PCs where we have access to a large screen and a full keyboard? So many questions we can and must address…

I pose a devil’s advocate question: But what if users are flocking to a cool new site, not already in the index? The answer: “A single page request can flag a new site to be included in the search index, giving unprecedented speed of index updating.”

But doesn’t that somehow at some level limit choice? Not really. To introduce an element of serendipity, the solution comes configured with a “throw external” feature that will redirect to GYM search engines if the results delivered by the operator search engine are below a given accuracy. (Settings for SafeSearch also allow the mobile operator to exclude harmful content from the search results.)

The source and I then debated the extent to which analytics and behavioral data could actually enable the creation of new and more personalized mobile search services. Our conclusion: It certainly covers all the bases to achieve this because it would only index the sites the subscribers like best. Take a mobile operator with a prosumer customer base such as O2. They would likely value straightforward results that help them plan and execute their daily routine, and the mobile search results would reflect this. A consumer-focused operator, such as Virgin Mobile, may find that its users prefer entertainment sites and social networking/moblogging destinations, and likewise give them top-notch placement in the mobile search results.

Beyond identifying sites to index tailored to the preferences of the subscriber base, the solution would pave the way for mobile operators to refine search results based on meta data from browsing activity. For the first time ever, mobile search results could be ranked by device type, time of day, or day of the week. As the source put it: “Initially we have  focused  on domain and page popularity within a given data set, however this can be further optimized, for example by refining result ranking based on  time of day or device – people  browsing in  the morning preferred a particular destination domain, while people with a Nokia N95 preferred another.”

Likewise, the solution would allow mobile advertising (which the industry is finally beginning to recognize is another form of content) to be refined and targeted to the preferences and profiles of the subscriber base. What’s more, the mobile operators could keep the lion’s share of the paid search revenues.

My Take: Operators – through their enviable position at the center of our mobile experiences – are perfectly positioned to tap into the wisdom of invisible crowds (us). With the right tools and technology mobile operators can follow our virtual breadcrumb trail to optimize our mobile search (and advertising) experiences, using our actual usage patterns to give us the answers/results we will most likely appreciate. What better way to compete against search engine giants and even win?

Tags: abphone, behavioral targeting, ChaCha, Mobile Advertising, mobile analytics, Mobile Marketing, Mobile Search, social search, Taptu, white paper

This entry was posted on Monday, March 16th, 2009 at 11:53 am and is filed under Content Discovery, Mobile Advertising, Mobile Research, Mobile Search. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.