Net apps tune up sales pitch

Three companies have rolled out products that they say will turn web-browsers into buyers based on information discovery and contextual commerce.

Companies want browsers to become buyers

New tools that present products to Web customers based on what Web pages they view and what they purchase will let e-tailers more precisely target their marketing pitches.

Merchandising Avenue Inc., Chiliad Inc. and Savage Beast Technologies Inc. have new technologies that they said will turn browsers into buyers.

Will these new app tools garner more sales? YES

Merchandising Avenue, a San Diego-based application service provider, this week will unveil its Merchandising Network. The hosted software ties into an e-commerce or a general-interest Web site and monitors what pages customers are viewing. Based on the content on those pages, the system dynamically adds a vertical banner, called Personal Shopper Recommends, to each page.

The banner contains an image, a small description and pricing for up to five products relevant to the content on a site, Merchandising Avenue officials said. The banner also offers shoppers buttons to purchase the products or get more information about them. Touching those buttons takes the customer to a Merchandising Avenue affiliate site that is selling that product.

A proprietary Smart Palette Serving technology uses JavaScript to determine the consumer's browser type and serve a compatible version of the Personal Shopper Recommendations. Behind the network are in-house merchandisers who read content and pick the top products in 50 or 60 categories related to it, officials said.

An e-tailer that has signed up as one of the initial 80 Merchandising Network affiliates looks forward to opening a new sales channel. Heidi Chu, affiliate marketing manager at Inc., a candle maker specializing in home décor, has limited expectations for how much revenue the Personal Shopper Recommendations will bring in, but because she spent very little time making her company's Web site and products ready to work with the network, she is happy to take part.

"I'd just like to be surprised in terms of how well Merchandising Avenue does with sales," said Chu, in Petaluma, Calif.

Merchants either share revenue or pay a negotiated percentage when consumers buy products. Alternatively, Merchandising Avenue will charge 75 cents per consumer click. Merchandising Avenue pays content sites half the sales commission received from merchants.

Separately, Chiliad, of Amherst, Mass., is developing software, due early next year, that will enhance the functionality of Web and wireless services to provide e-merchants with precise information discovery and to enable contextual commerce and Internet advertising, officials said. The company this month received $24.5 million in financing from Hewlett-Packard Co.

Meanwhile, startup Savage Beast is looking to provide context to content recommendations made on music retail Web sites. The Oakland, Calif., company's Music Genome Project, which was launched last week, is a database of 250,000 songs, analyzed and categorized across 180 musical, lyrical and compositional characteristics.

In addition, Savage Beast is offering Music Hunter, which matches songs that customers know to songs from the database that they might also be interested in. A second tool, called Profiler, is a personalization engine that makes recommendations based on customers' tastes. Recommendations can be scored based on criteria that customers select, such as the closest match for a brass solo or a funky groove.

The technology is designed to help music sites introduce customers to artists they might not be that familiar with, thereby increasing sales of lesser-known artists, a big deal for an industry that relies heavily on best-selling artists for most of its profits., a site for independent recording artists, is an early adopter of the Savage Beast technology. Riffage will incorporate about one-third of its 30,000-song database into the Music Genome Project on its site. Customers can enter song titles they know, and the Match Hunter will find songs with similar characteristics on Riffage.

"If you like Beck or Phish, you can find artists on Riffage whose songs most closely resemble theirs," said Scott Maddux, director of product development at Riffage, in Palo Alto, Calif. "Most people base their music preferences on music they already know. That's what makes this such an intuitive and powerful tool."

Maddux likes that Savage Beast relies on professional musicians to analyze music for similarities, rather than machine-based systems that analyze only sound-wave patterns.

"The machine approach is faster, but it can't analyze lyrical content. It can't even tell if it's a male or female vocalist," he said.

Maddux hopes that Savage Beast's technology will increase the exposure of artists on Riffage. Artists who are successful on Riffage can get booked at the company's 600-seat concert venue in San Francisco, the Great American Music Hall, and signed to its record label, 1500 Records.

"We have a music ecosystem, and the site is the farm," Maddux said. "There's a lot of great music on our site by independent artists that most people have never heard of."

Music Genome Project pricing varies, depending on the customer site. Although pricing is generally performance-based, Savage Beast officials said they may also charge upfront fees for licensing, enrollment and song library analysis.

Savage Beast CEO Jon Kraft has plans to take the technology beyond music. He's acquired the Web domains for "Movie Genome Project" and "Book Genome Project" as well.