Easy Does It
By Peggy Anne Salz, Mon Apr 18 08:15:00 GMT 2005

Navigating to content through a series of menus can be frustrating for users. Recommendation engines promise to remove the pain from the content discovery process, and take providers' sales to new levels.

Give users what they want? Not if they can't find it.

An industry rule of thumb from the fixed Internet was content providers lose half their audience to frustration or ennui at every additional click it takes them to find what they want. Why should the mobile Internet be any different?

Indeed, a usability report from Norman Nielsen Group found that, to be genuinely useful, mobile portals must be able to bring content to users within approximately 30 seconds. Put another way, users are unlikely to discover content that is more than 30 seconds from the portal homepage. While the research is a few years old, its message was never more relevant.

More recent insights into the ideal user experience come from ChangingWorlds, an Irish provider of artificial intelligence solutions. Together with Mobile Metrix, a Swedish research & consulting firm, ChangingWorlds evaluated the degree of effort users have to expend to locate and access content they like across 19 European mobile portals.
It found that users consider 12 clicks to be an acceptable click-distance (the number of clicks/menus to arrive at the desired content). Given this threshold, a whopping 65% of mobile content and applications are positioned too far away from the portal homepage to attract users.

An Offer You Can't Refuse

Operators are now looking to implement recommendation engines modelled on the Amazon approach ("people who liked this also liked this") to drive data usage and allow users to circumvent the mobile interface problem by connecting directly with relevant content their peers recommend.

Against this backdrop, ChangingWorlds has "harnessed artificial intelligence to generate recommendations based on the content both users themselves and other users in the community have looked at or downloaded," explains product marketing manager John Doyle. "Because the rich user profile is based on what users do rather than what they say, and doesn't require user input of any kind, the information is always accurate and up-to-date."

Qpass, a content platform provider with more than 500,000 digital content assets from over 1,000 content providers, is also working on a recommendation engine it plans to make commercial this year. It believes its insight into over 150 million transactions per month allows it to make more credible recommendations about what the majority of users worldwide like.

But knowing what users prefer is just part of the picture. For this reason, solutions that simply attempt to transplant the Amazon approach to the mobile space will fall short, argues Franz Jachim, Qpass' vice president of technology, Europe.

"You can't just recommend content; you have to know that when you recommend an MP3 file from a similar artist that the user has a phone capable of playing the file in the first place. Otherwise, the tip is useless and the user is more frustrated than before," Jachim explains, so Qpass' solution will include this kind of customer data.

A Trusted Source

But some companies believe that users are more likely to respond to recommendations from users they actually know, or can at least identify. This is the logic behind solutions from Ireland's NewBay. Its FoneShare product provides users with a "personal space" in the mobile Internet and allows users to set up and join communities of like-minded people. "We provide an addictive, interactive zone for users to find, buy and recommend cool content while making new friends," explains John Barron, a NewBay manager.

Put simply, FoneShare encourages users to recommend content to other users with similar interests. It supports this process by enabling users to create and join mobile clubs around their interests and chat with others who like the same things. Community members earn points for content recommendations that result in a sale, and users can also cash in these points to receive additional content.

It's clearly in operators' interests to deliver effective and targeted commerce experiences to their customers. This means delivering the right content to the right users. Recommendations based on generic information such as page views and downloads will be an important part of this strategy, but it may well be the recommendations from the tight-knit communities users know and trust that clinch the sale.