Advertisement

Getting Bizzy with e-commerce preference engines

Share

This article was originally on a blog post platform and may be missing photos, graphics or links. See About archive blog posts.

For a digital music or movie service, a preference engine is like a car stereo. It’s not indispensable, really, but the service is a lot less appealing without one. And some, like Netflix’s, are so good at recommending items that they become a central feature of the user interface.

So it’s not surprising that companies are starting to bring similar technologies to other types of e-commerce. The main question is what data to use to generate recommendations -- whether to rely on user-generated reviews, a person’s Facebook friends or some other method for matching people to products.

Advertisement

Mountain View, Calif.-based Bizzy, a start-up that launched on the Web last month, is taking a page out of Netflix’s book for its service, which recommends restaurants, stores, nightspots and a variety of other outlets and services in the user’s community. Bizzy asks new users 20 questions about their favorite places, creating a profile that it compares to other users’. It then draws recommendations for them based on the favorites entered by the people whose tastes are closest to theirs.

There’s plenty of competition in the market, most notably from Yelp, which collects and displays reviews of local businesses. But Yelp shows the same listings to anyone who searches for, say, restaurants in Los Angeles, while Bizzy returns a personalized set of recommendations that varies from user to user.

Nor does Bizzy try to derive any insights about people from what their friends say they like, unlike Google’s new HotPot service.

Instead, Gadi Shamia, Bizzy’s founder and president said, the company looks across its user base in any community to find people whose list of favorites match in any given category -- or even sub-category. ‘If you and I are similar in five out of 10 [favorite] restaurants, but four of them are coffee shops, my recommendation on coffee is going to be most important to you,’ Shamia said. Ryan Kuder, Bizzy’s vice president of marketing, contended that a person’s social graph just isn’t that useful in a preference engine. Kuder noted that he and Shamia are Facebook friends and ‘virtually identical’ by most of the demographic metrics used by marketers, such as income, family size and neighborhood. Yet their tastes differ sharply, he said, adding, ‘I don’t trust his restaurant recommendations.’

The quality and range of Bizzy’s service should improve as its user base grows and offers more favorites; my sampling of the recommendations for my town (South Pasadena) indicated that there aren’t enough users there yet to produce compelling results. To help build the Bizzy community, the company introduced a free iPhone app Tuesday that it hopes will help the service spread virally.

The questions Bizzy asks also focus on a younger, less family-oriented demographic than the one I occupy. But that makes sense -- no one in my demographic needs to be told where to look for food (the clearance section at Vons), books (the county library) or booster seats (Target).

Advertisement

Analyst Mike McGuire of Gartner, a market research firm, said it’s going to be interesting to see how well Bizzy and its competitors apply what they learn about users’ preferences to a wide variety of local offerings, and whether they can meld their recommendations with services that drive demand, such as Groupon.

Shamia said the company was focused on only one thing at the moment: ‘giving as many people as possible the best recommendations.’ Looking ahead, though, he said he sees ‘almost unlimited’ ways to monetize Bizzy’s audience. Particularly promising are deals that would use Bizzy to introduce businesses to the customers who’d be most likely to be interested in them. For example, he said, companies that aren’t at the top of the list of recommendations might want to offer coupons or freebies, with Bizzy being paid if the coupons are redeemed.

Related:

The Echo Nest: a powerful preference engine

-- Jon Healey

Healey writes editorials for The Times’ Opinion Manufacturing Division.

Advertisement