If you’re like me, you do a lot of research online before you make a big buying decision, and sometimes even before a small one. And you lean heavily on the reviews left by other shoppers. My favorite trick, when I’m zeroing in on a specific product, is to read all of the “1 star” or “most critical” reviews first. It’s a surefire way to find out about product glitches, misleading advertising claims, or poor customer service.

In fact, I’d say that the explosion of user-contributed reviews around the Web is one of the three biggest developments benefiting consumers in the e-commerce era (which dates back to mid-1995 or so, when Amazon first turned on its website). The other two are the rise of comparison shopping, which keeps prices in check, and the simple-but-still-amazing fact that you can click a “Buy” button on a website and make a book or a game or a DVD appear on your doorstep without ever leaving home. If you think about it, though, e-retailers themselves haven’t innovated that much around reviews. A review on Amazon today looks and acts pretty much like a review on Amazon in 2002.

But there’s one San Francisco company, PowerReviews, that keeps working to make reviews more useful for consumers and more profitable for e-retailers. I spent some time recently with Andy Chen, PowerReviews’ founder, and learned a lot about how the company sees the world of product research—and what it’s learning that could change the experience of shopping online.

Andy Chen, PowerReviews Founder

Chen says that when the company got started back in 2005, it had two interlinked ideas. The first was that a lot more e-retailers needed customer reviews on their sites, and that it might be profitable to build a white-label review system that companies could plug into their sites rather than having to build the functionality on their own. The second idea was that if you could build this system, it would then be possible to aggregate all of the reviews collected by all of the participating e-retailers into one giant product recommendations site, and use that site to earn money on lead generation and affiliate commissions. That’s exactly what PowerReviews has done with its sister site Buzzillions, which brings together a good chunk of the 23 million reviews sent in by customers of Power Review’s clients. That, by the way, is a group of almost 6,000 companies, including many high-profile bricks-and-mortar brands like Toys R Us, Staples, and REI.

PowerReviews has had to tweak its ideas along the way. It originally provided the review system to e-retailers for free, but the Buzzillions site didn’t turn out to be as profitable as the startup had hoped. So the company had to start charging for the review platform, just as competitors like Austin, TX-based BazaarVoice do. The good news is that “we’ve been growing extremely quickly” on the platform side, while Buzzillions is “still a good revenue source,” according to Chen, whose formal title is Founder and Vice president of Strategic Partnerships. (Pehr Luedtke, a veteran of Shopping.com and Levi Strauss, is PowerReviews’ CEO.)

With some $37 million in venture capital on the line from Menlo Ventures, Tenaya Capital, Draper Richards, Four Rivers Group, and Woodside Fund, PowerReviews needs all the growth that it can get. To speed things up, and to compete with e-commerce juggernaut Amazon, the startup has been trying new strategies like getting consumers to answer each other’s questions about products (that’s a feature called AnswerBox) and integrating the review platform with social media, especially Facebook.

But one of the company’s biggest opportunities, Chen says, may not lie in the reviews, exactly, but rather in what he calls “social navigation” based on all the data PowerReviews collects about the reviews and the reviewers. By analyzing what products people are … Next Page »

Comments (1) | Reprints | Share:          

Please…not yet another recommendation engine for telling me where I should eat, what clothes to buy, or which places to visit on my trips. Didn’t StyleFeeder get acquired by Time Inc. last year? Didn’t Where just get bought by PayPal? But I guess that’s why startups keep trying to innovate with recommendations—if they do it right, there will be a big payout coming.

OK, well, having a cute fish logo doesn’t hurt either. Last Tuesday, Boston-based Goby released a new, overhauled version of its iPhone app, and it represents an important shift for the three-year-old startup. That’s right—Goby, which started in 2008 as a “deep Web” search engine focused on helping consumers find events and activities near them—is now going hard after recommendations.

“It’s a big shift away from search and browse,” says Mark Watkins, the company’s co-founder and CEO.

Goby already had a mobile app on iPhone and Android devices, but the new app’s emphasis is on understanding consumers’ preferences and showing them a short list of relevant activities, instead of just helping them search through long lists. (Goby made the change in part because of user feedback, but you can still search in the app as well.) Before last week’s release, Watkins said the company had some 750,000 mobile-app downloads, and that its user base was split about 50-50 between mobile and Web. Reached by e-mail yesterday, Watkins says the new app had about 70,000 downloads in the past week, and that about half of the new users have created personalized “fun feeds” (more on that in a minute). Overall, not a bad start for the new Goby app.

Mobile has been a big focus at the 10-person startup over the past nine months. Goby, along with a lot of other companies, seems to have adapted quickly to the reality that mobile is a better business opportunity for its technology than the Web. “People are hungry for mobile apps, and advertisers are hungry to reach them,” Watkins says.

The new app is all about showing people five to 10 activities that interest them, Watkins says. Goby scans users’ Facebook profiles and uses semantic text analysis to create a “topic map” of things people like. The categories include sporting events, live music, museums, food and drink, and so forth. Goby also has access to a huge database of events and activities, culled from a variety of sites around the Web. By personalizing your own feed, you can do things like get alerts for heavy-metal acts playing in Worcester, MA, or jazz artists coming through Boston. The app also makes it easy to share the info with your social network through Facebook and Twitter, Watkins says.

The technology sounds simple—but it’s not. One issue is that semantic analysis and … Next Page »

Comments (1) | Reprints | Share:          

Send the Trend, a New York-based e-commerce startup focused on personalized fashion accessory recommendations, has raised $2.5 million of an equity-based funding round that could hit $3 million, an SEC filing shows. The money comes from two investors, according to the document. Send the Trend’s website names Christian Siriano, a previous winner of the television show Project Runway, as a chief curator and stylist for the company.

Comments | Reprints | Share:          

Variety was the theme of the Northwest Entrepreneur Network’s First Look Forum yesterday at the Arctic Club Hotel, a venerable establishment in downtown Seattle. There were a dozen startup companies with a wide array of ideas, each giving a five-minute pitch to an eclectic audience that included investors, entrepreneurship coaches, media, and sponsors. The startups were there to compete for a chance to win some cool prizes, and of course, score new investments. The “first look” aspect means none had presented previously to investors or venture capital groups—this is the brainchild of NWEN executive director Rebecca Lovell.

One interesting point: the initial group of companies represented much more than just software and tech. The bunch covered medical devices, techniques for combating everything from obesity to tooth decay to slow and messy colonoscopies, and applications across industries ranging from online travel to trucking to pet care. Only three of the 12 startups in the backyard of Microsoft were software companies.

But in the end, the five finalists—selected by audience voting—consisted of those three software startups plus two hardware/materials tech companies (though one of those has a biomedical application). Any of these certainly could make a promising business, but it’s also possible that the investors and other voters in the room were just more comfortable with tech than life sciences or medical companies. Or maybe it’s just that tech startups tend to require smaller amounts of startup capital that can lead to higher potential returns.

Here are my initial impressions and super-short summaries of what each of the 12 Northwest companies presented (plus more details on the finalists and winner below):

Biomoles
Developer of higher quality and purity techniques for doing DNA purification for automated DNA processing and life sciences applications.

Crux Medical Innovations
Producer of a special biopsy device that makes procedures like colonoscopies (done 20 million times a year in the U.S. to detect colon cancer) faster and more efficient.

Darwin’s Natural Pet Products
A six-year-old company that makes and delivers fresh, healthy frozen meals for cats and dogs; it has more than 1,500 customers, mostly in the Seattle area, and wants to go national.

DragGone Aerodynamics
Maker of add-ons to trucks that reduce fuel consumption by improving aerodynamics.

Empowering Engineering Technologies
Three-year-old company developing an elastic tendon-like medical device that helps people with gait disabilities walk; the orthopedic technology comes Cleveland Clinic and is currently being tested in patients. (This company is also presenting Wednesday at the inaugural meeting of Wings, the Northwest’s new medical device angel network, as Luke reported yesterday.)

HealthyLogics
A startup developing “biofeedback” utensils, like a special fork that measures how fast … Next Page »

Comments (4) | Reprints | Share:          

The next big startup out of Seattle might have been conceived on the top floor of the Washington Athletic Club this past Tuesday. That’s where 35 carefully selected tech entrepreneurs and business people gathered over drinks and light appetizers for the first-ever FounderDating session in Seattle.

The idea, as we reported on a couple weeks ago, was to have an event dedicated to helping prospective co-founders meet new people who could be complementary in starting companies or at least talking about ideas. FounderDating started in the San Francisco Bay Area, and Brian Schultz—formerly of Seattle startups Ontela and Djinnisys, and now at Microsoft—helped bring it to the Northwest.

I asked Schultz yesterday how the event went. He says he is in the midst of surveying the attendees, who were culled from more than 100 applicants. “I’m really excited about the feedback,” Schultz says. “People were definitely passionate about it.”

The tech sectors represented that evening included mobile, enterprise software, advertising, commerce, and recommendation systems. Nametags helped distinguish business people from tech people (engineers), as well as their sectors of interest. Schultz says there was a big range in experience—some people were in their first jobs out of school, while others were former CEOs with multiple successful startup exits or executives at big companies.

The identities of the attendees are supposed to be kept secret to the outside world, because some have employers who might not appreciate hearing that they’re looking to start something new. As Schultz jokes, the first rule of FounderDating is you don’t talk about FounderDating. (Which makes me think, a CEO fight club in Seattle would be pretty awesome. I can think of some good matchups already.)

There was freeform mingling, a 60-second speed dating round, small-group activities, and more mingling. Early reviews seem to indicate most attendees weren’t recruiting for a specific startup, but were focused on meeting a variety of new people. Several wanted to see more younger and hungrier engineers. “It’s really about the people,” Schultz says—not necessarily their startup ideas.

In any case, it sounds like the event was successful enough that Schultz will try to organize them quarterly in Seattle, and create stronger linkages between Seattle participants and those in the Bay Area via the FounderDating website. The next Seattle event will probably be in May or June, Schultz says, and he’s looking for more help in organizing and screening applicants.

Schultz also says he wants to survey the same 35 people three to six months from now, and see what new projects have come about from their discussions. Because ultimately that’s what this is about—doing something new to help startups get formed. “Can you attribute anything to the people you met that night?” he says. “There were definitely some new connections made.”

Comments (3) | Reprints | Share:          

[Updated, see below] Time Inc., the New York-based magazine publishing giant, has acquired Cambridge, MA-based personalized shopping site StyleFeeder in a deal that closed on Friday. StyleFeeder founder and chief technology officer Phil Jacob, who I reached at his home this evening, confirmed the news, which was first publicly reported just before 8:00 p.m. Monday by the Wall Street Journal.

An unnamed source in the WSJ report pegged the purchase price at an amount “well into eight figures.” Xconomy has not been able to confirm that claim, and Jacob said he was unable to comment on it.

Four-year-old StyleFeeder, which was backed by Lexington, MA-based Highland Capital Partners and Boston-based Schooner Capital, provides personalized shopping recommendations to visitors using a combination of techniques. For users who are logged into the site, and for whom StyleFeeder has some historical data, the company uses machine learning algorithms developed at MIT to recommend products rated highly by other users with similar tastes. For users who aren’t logged in, the company can still make useful recommendations based on users’ locations and records of products that are popular in their areas.

Last March, StyleFeeder branched out, announcing a partnership with Hachette Filipachi Media, the publisher of Elle and a number of other fashion, automotive, health, and hobbyist magazines, under which StyleFeeder agreed to provide personalized product recommendations directly to Elle.com visitors. The affiliate fees that accrue from such purchases are becoming an attractive revenue source for many consumer-oriented publications. Fran Hauser, a digital strategy manager for several Time Inc. magazines such as People and InStyle, told the Wall Street Journal that with the acquisition of StyleFeeder, Time Inc. will be able to “share in [the] value creation” when editors help to generate consumer demand for products.

It’s safe to say, in other words, that online product finders like this one that StyleFeeder created for Elle are likely to turn up on the Websites of many fashion-oriented Time titles, including InStyle. Whether StyleFeeder will be allowed to continue its partnership with Hachette Filipachi and other brands outside the Time Inc. empire remains an open question.

The WSJ report said an official announcement of the acquisition could come Tuesday. [Update, 8:30 a.m., 1/19/10: Time Inc. has just published a press release announcing the news. It indicates that StyleFeeder "will become part of the InStyle family of brands" and that it will power shopping on InStyle.com.]

Thanks to growing affiliate revenues and conservative management, StyleFeeder has been a cash-flow-positive business for some time, and had a strong holiday season. According to a source familiar with the company, the startup did not need to be acquired to stay in business, and was not actively seeking a corporate parent. “Somebody came along,” this source tells Xconomy.

That somebody was clearly Time Inc., the magazine group of media conglomerate Time Warner. In addition to People and InStyle, Time Inc.’s titles include magazines such as Fortune, Sports Illustrated, Health, Real Simple, Money, This Old House, Golf, and of course the flagship Time. The company claims that one out of every seven U.S. Internet users visits a Time Inc. website at least once a month. It has been actively exploring new types of online transactions such as digital magazine sales that could help to offset declining advertising revenues at its big-name magazines.

As a venture-funded startup, of course, StyleFeeder would eventually have needed to achieve some sort of exit. The company has raised $4 million in seed and Series A funding from Highland and Schooner. If the Wall Street Journal‘s “eight figures” quote is correct—meaning Time has paid more than $10 million but less than $100 million—the acquisition represents a return of between 2.5x and 25x on the venture funds’ investment.

Comments (2) | Reprints | Share:          

Andover, MA-based envIO Networks has spent several years developing a recommendation technology called “Social Genome” that helps mobile phone owners find applications they’ll like based on the preferences of their friends and acquaintances. Today envIO introduced software called Chorus that applies the Social Genome technology to the 100,000-plus apps in the Apple iTunes App Store. Chorus, which is itself an iPhone app, lets users see which other apps their friends are using and how they’re rated. “iPhone users already ask friends and colleagues about cool, new apps. Chorus makes it simple,” Linda Barrabee, a former Yankee Media and Jupiter mobile analyst who is an advisor to envIO, said in a statement.

Comments | Reprints | Share:          

Yesterday we published Part 1 of our interview with Philip Jacob, founder of Cambridge, MA-based StyleFeeder, and Shergul Arshad, the startup’s vice president of business development. The two men talked about the remarkable growth of the personalized shopping site, which has taken in $3.5 million in venture capital since 2007 and is already profitable. In Part 2 below, Jacobs and Arshad share their changing perspectives on Facebook (where StyleFeeder has a popular third-party shopping application), the evolution of the company’s machine-learning technology, and the different ways StyleFeeder caters to male and female audiences.

Xconomy: You said you had a million registered users. Does that include all the people who access StyleFeeder from inside Facebook?

Shergul Arshad: No, if you count Facebook the number is much higher. 2.5 million people have installed the StyleFeeder app on Facebook. But usership goes up and down depending on the latest bug or redesign that Facebook is going through. It’s not fair to count all those.

X: How important is Facebook for you these days as a channel? At one time it was your main source of new members, wasn’t it?

Phil Jacob: It’s no longer what it once was. Facebook has really deemphasized applications through a series of user-interface changes that make it impossible at this point to scale to the level that a Slide or a RockYou got to. We still have people using it, and there is this viral nature where people can see what other people have posted to their Facebook feed, and it does drive some traffic, but the window of opportunity for third-party apps is gone. But it was a huge success for us, in terms of driving awareness and registered users. It was the single best thing we’ve done from a growth standpoint.

X: What are your thoughts about the direction Facebook has taken?

PJ: It’s interesting. They have this “verified app” program, and we just got rejected, even though we are the largest shopping app on Facebook, because we have clothes on our site, and one of the pieces of clothing we have are thongs. Out of 14 million products, that is the one that got us rejected.

So Facebook is in a weird situation. I think they are struggling for a number of reasons. Their capital costs are just crazy—at one point they were spending $2 million a week just on new storage for their photos…I don’t want to get into an East Coast versus West Coast thing, but I think there is also a difference in mentality that occurs when you actually start focusing on revenue. Everybody can get aligned behind that; we can have conversations with investors and the conversations are so much easier because they’re not asking who are you, what are you—the numbers are there. But if you keep changing your goal as time goes on, there’s always some other metric that seems to be the important one. Facebook seems to keep shifting, which I think is dangerous for them.

X: Their most recent redesign seems to have been conceived to make them more like Twitter—which, ironically, is another West Coast startup without a revenue focus.

SA: It was publicized that they at one point wanted to buy Twitter, and it seems like somebody pretty high up within Facebook said, “If we can’t buy them, we can clobber them at their own game.” I think, personally, that has caused a lot of backlash. People use the tools in different ways. I used to benefit from both Facebook and Twitter greatly, but I think Facebook is starting to be in a position where by constantly prompting people for status updates, you just get more and more inane banter. By suppressing the third-party apps, Facebook is in many ways biting the hand that was feeding them. With an app like poker, a lot of college-aged people can play poker all day, and that was driving page views and ad sales, which is their core revenue model. But if you suppress poker in favor of inane banter, you’ve put the page views toward banter creation versus extracting value.

X: At the beginning, a big part of StyleFeeder’s story was around the machine-learning algorithms that Jason Rennie from MIT built to power your recommendation engine. Have you continued to refine and tweak that engine?

PJ: We don’t make dramatic changes to it. It works quite well the way it’s built. What I’m more interested in—and we have a big effort ongoing right now, with a product launch coming in another six weeks or so—is understanding more about the products we have on the site. We have thousands of data feeds coming to us, and as you might imagine, there is a varying level of data quality between Amazon at one end and … Next Page »

Comments (1) | Reprints | Share:          

StyleFeeder is that rarest of animals: a profitable Web 2.0 company, and on the East Coast to boot. It may also have one of the largest ratios of users to employees of any Web business: more than a million registered members, served by a grand total of only six full-time employees. I met all six a couple of weeks ago when I stopped by the company’s offices in hip Central Square, Cambridge, MA, for a progress report.

As I explained in a January 2008 profile, StyleFeeder offers a shopping search engine that uses sophisticated machine-learning algorithms to study users’ tastes and recommend new products from thousands of online retailers. Anytime a user clicks on a recommended product and ends up buying it, StyleFeeder earns a nifty commission. Recently, the startup rolled out a personalized shopping service for the Elle magazine website, and plans to expand the partnership with Elle owner Hachette Filipacchi Media to the publisher’s other websites, which cover the fashion, automotive, luxury design, women’s health, and hobbyist markets.

Software engineer Philip Jacobs started StyleFeeder in late 2005, sold it to now-defunct blog aggregator TopTenSources in 2006, and spun it out again in early 2007. The company has raised $3.5 million from Highland Capital Partners and Schooner Capital—a modest amount, as tech startups go. But it’s been enough to power the company all the way to profitability, thanks to some shrewd decisions that have kept the company’s costs low.

One, as Jacobs explained at Xconomy’s cloud computing forum last year, was to outsource the company’s entire IT infrastructure to Amazon Web Services. Another was to keep the staff small and hire freelancers to do much of the coding scut work. A third was to jump on the Facebook bandwagon at exactly the right time—-mid-2007, when the social networking site was opening the doors to providers of third-party applications. StyleFeeder’s intrinsically social approach to personalized product recommendations vaulted it to the number-one spot among Facebook shopping apps, which brought hundreds of thousands of new members streaming into StyleFeeder itself.

As Jacobs and Stylefeeder’s vice president of business development Shergul Arshad told me during my recent visit, the Facebook wave has largely crashed and dissipated for StyleFeeder (and for a lot of other third-party app providers). But that hasn’t slowed the company’s growth—a record 2.1 million unique visitors stopped at the site in March. I talked with Jacobs and Arshad about how they’ve managed that expansion, how the company’s relationship with Facebook has evolved, where its recommendation technology is going, and how women and men use the site differently.

As you’ll see, they didn’t need a lot of prompting from me to wax on about the company’s business model, which clearly still energizes them, even as the company approaches its fourth birthday. Part 1 of the lengthy interview appears below; we’ll publish Part 2 on Wednesday. [Update 4/29/09: Part 2 now published.]

Xconomy: It’s been a while since we talked about StyleFeeder. How have things been going?

Philip Jacob: We’re kind of in this interesting space. The economy has not done well, but we’re not only doing well, we’ve been doing really well for a long time. At this point, we’re profitable, which is unusual not only for a Web 2.0 company but for an East Coast startup.

We learn about people’s tastes and preferences and help them with their online shopping. If you say you like jeans, we know what kind you might like based on your other preferences. We have this tricked-up collaborative filtering system that powers personalized search and production recommendations and will also introduce you to other people who shop like you do. We are continuing to execute on that and adding new features and monetizing the site through affiliate relationships and our partnership with Hachette Filipacchi. We’ve been up live on the Elle Magazine site for about five or six weeks now.

We’ve done a lot with a little, which was an idea that was part of the original wave of Web 2.0 startups that sort of got lost. People started out taking $2 million or $3 million [in venture capital], which was great. But then when you start taking $10 million or $20 million, that starts to look like … Next Page »

Comments (1) | Reprints | Share:          

Want to ask your network of friends and colleagues for advice on where to stay when you’re traveling? Or maybe you want to offer recommendations for a local electrician, dentist, or specialty food shop. You could do it all by e-mail, or through existing social networks like Facebook. But dealing with such requests and recommendations as they come up can be clumsy and inefficient—who needs more random e-mail threads cluttering up their inbox?

FriendlyFavor claims to have a better solution. The Seattle startup is rolling out a national website today that lets you keep all your social requests and offers in one centralized place. The site works as a request tool for everything from job referrals to consumer recommendations—directed to your specific network of people whom you know and trust. The software is being designed to hook into social-networking sites like Facebook, MySpace, and LinkedIn.

And that’s the key to FriendlyFavor’s strategy. “It’s not another social network,” says president and co-founder Scott Larson, who previously founded Vashon Partners. “It’s an application to complement those networks you’ve already built…in a trusted and more purposeful way.” Larson says its value lies in being more efficient than e-mail in managing requests and favors.

Beyond marketing the tool to consumers, a next step for FriendlyFavor will be to license the software platform to businesses and organizations, says co-founder and chief strategist John Patton. That means university alumni networks, marketing teams in large companies, and so forth. Patton emphasizes the importance of having “a single location for each favor” that’s persistent, as compared to the transient messages on Facebook, for example.

FriendlyFavor was founded in 2007 and has raised $500,000 in angel funding. It is backed by individual investors from companies like Microsoft, Amazon, Google, and Marchex. FriendlyFavor currently has five employees, and it launched a public beta trial of 500 users in the Seattle area last summer. The startup generates revenues from a combination of targeted online advertising, technology licensing fees, and sponsorships.

Comments | Reprints | Share:          

There’s a startup in Cambridge, MA, that’s quietly collecting gigabytes of data about the likes, dislikes, habits, and exploits of hard-core gamers. Given how lucrative console and online gaming have become—bringing in $1.3 billion for U.S. publishers in October alone, up 18 percent from the same month last year—you might guess that the startup intends to sell its insights to game companies, perhaps as a way of targeting gamers with advertisements and offers.

But that’s not the idea at all. The company, GamerDNA, was founded by and for gamers—and next week it will formally unveil the piece of software that’s making use of all of that data, a “discovery engine” designed to help gamers find new games they might like and connect with other gamers who have similar tastes.

“At GamerDNA the consumer is number one in everything we do,” says founder and CEO Jon Radoff, whom I’ve interviewed a couple of times before. “It’s all about helping you find other people like you, games you’ll enjoy, communities where you are going to feel at home, and tools for creating your identity as a gamer online.”

The discovery engine, which has been in a beta testing phase for several months, is a more sophisticated alternative to other automated game-recommendation systems, which are pretty much limited to the Amazon-style “customers who bought Final Fantasy XII also bought Kingdom Hearts 2″ genre. In other media, such as music, fans and developers are outgrowing the old-fashioned collaborative filtering technology inside most recommendation systems (including Amazon’s), and are looking for fresh information sources that cater to their unique personal tastes. That’s what GamerDNA’s system is designed to do for the console video game and online role-playing genres.

The engine works in part by breaking down user feedback into categories: setting, tone, who the user plays as, who the user plays against, and how the game works. “For some people, it’s all about the world,” explains Radoff. “For others, it’s who you get to be in that world. People want to be able to search along those particular lines, because they’re going to get a result that’s more relevant to them.”

During a visit to GamerDNA’s groovy Central Square office last week—which must be right above an Indian restaurant, since I left with a huge craving for chicken makhani—Radoff walked me through the discovery engine’s process for plumbing what users like about various game, and then serving up informed suggestions about what other games they might like.

I started by entering the title Bioshock, one of my favorite Xbox 360 games, into the GamerDNA discovery engine’s search box—the same way a music fan might enter an artist or a song name into the search box at Pandora’s personalized streaming music service. Then the system asked me, “Why do you like Bioshock?” Well, to me, the coolest thing about the game is … Next Page »

Comments | Reprints | Share:          

At least two of us here at Xconomy—Rebecca and myself—are huge fans of Pandora, the Oakland, CA-based streaming music company. My enthusiasm has only grown since July, when Pandora released an iPhone application that, I think many users would agree, is the single most useful and enjoyable third-party app available for the device. (It’s currently #30 on the “Top Free Apps” list at the iTunes App Store, but I think that’s only because everybody has already downloaded it.)

So when I heard that Tim Westergren, Pandora’s founder and chief strategy officer, would be speaking at the Apple Store on Boylston Street in Boston on Monday, I made like any other self-respecting Pandora fanboy and cleared my schedule. The table-full of listeners who joined me to hear Westergren’s talk starting at noon Monday gradually grew to a crowd of perhaps 50 onlookers.

This being the Apple Store, Westergren spent part of his time explaining how the Pandora iPhone app works. But he also gave an interesting look behind the scenes at Pandora, explaining how the company’s musicologists analyze songs, how the company itself survived nearly three years on the edge of extinction, how it monetizes its free streaming service today, and how it dealt with the threat of changes in performance-royalty rates that threatened once again this year to put the company out of business.

In a nutshell, Pandora lets users create personalized, streaming radio “stations” that play tunes by artists they like, along with music from other artists with similar styles. Its system decides what to play based on the “proximity” of one song to every other song in the company’s database, where songs are scored based on hundreds of attributes such as genre, key, tempo, the gender of the lead vocalist, and the timbre of the various instruments. Users can give a thumbs-up or a thumbs-down to every song that’s played; Pandora’s algorithms use this information to adjust the playlist for each station over time.

The iPhone version of Pandora—which wasn’t possible until Apple allowed third-party apps onto the device last summer, and wouldn’t have been practical, in any case, until the broadband iPhone 3G came along—does essentially everything that the Web-based version does. It’s by no means the only mobile version of Pandora—the company has actually built about 60 different versions of its software for various phones sold by Sprint and AT&T, and is also working on a version of the app that will run on RIM Blackberry devices—but the iPhone version is the first one that has really demonstrated the potential of the Pandora service on an untethered device, Westergren says.

“The iPhone has really changed the trajectory of the company,” he says. Since July 11, the day Apple launched the iPhone 3G and the iTunes App Store, 1.8 million iPhone and iPod Touch users have downloaded the Pandora application, and people listening on these devices are now consuming 10 percent of all music streamed by the company, he says. And whereas the company’s peak listening times have traditionally been 9:00 a.m. to 5:00 p.m.—when people are at their desks at work—that period has stretched out by an hour on either side of the work day, as iPhone owners tune into Pandora during their commutes.

(As for Android, the Google-led, open-source phone operating system, Westergren says it’s still “up in the air” whether Pandora will build an app that works on that platform. “We want to see how much it gets adopted first,” he says.)

Westergren shared some hopes for the mobile versions of Pandora that will make owners of commercial broadcast radio stations either shiver or snicker. “Our objective is to completely replace music radio with this,” he says. And that includes listening to music in your car. With a cheap adapter, Westergren notes, it’s easy to stream Pandora music from an iPhone to a car stereo. “I encourage you to try that,” he says. “It’s weird—I hadn’t actually used it in the car until a few months ago, and one day I was driving along and I said, ‘Hey, … Next Page »

Comments (1) | Reprints | Share:          

As I walked through a windy, chilly Somerville on my way to visit music-discovery startup The Echo Nest yesterday, last weekend’s sudden shutdown at Matchmine weighed on my mind. After all, The Echo Nest is in the same general business as the now-defunct Needham, MA, company: building software that helps digital media companies provide Web users with personalized recommendations, by rating the intrinsic qualities of songs or other media files in the companies’ catalogs and matching them with the tastes expressed by users. And both companies were built around the belief that today’s dominant recommendation systems, based on a decade-old technique called collaborative filtering, do a lousy job of helping people discover new music and other media.

But after pouring nearly two years of work and $10 million in financing into its MatchKey service, Matchmine was unable to find enough clients to keep its main investor, the Kraft Group, from pulling the plug in a bid to help stem its own cash flow crisis. So I had lots of questions for the guys at The Echo Nest—who just brought their company out of stealth mode last month—about whether there’s a real demand for new music recommendation technologies.

As it turns out, they had some pretty good answers. While the company hasn’t said which media companies it’s working with, CEO Jim Lucchese says the startup has been approached by a long list of music-driven organizations—think social networking sites, Internet radio stations, and the like—that are racing to upgrade their recommendation services to compete with hot online music platforms like Pandora and Last.fm. “Recommendation has moved from a nice-to-have to a must-have,” Lucchese says. “There are a lot of companies that need to get into parity with their competitors.” The Echo Nest plans to start talking about some of its customers in a couple of months, Lucchese says, and by the end of the first quarter of 2009, he predicts, “we’ll be powering applications on a number of the comScore top 10 music properties,” referring to websites monitored by the audience measurement service comScore Inc.

An MIT spinoff founded in 2005 (nearly two years before Matchmine), The Echo Nest may also be benefiting from an early decision to stay small and raise little capital. That meant the company could afford to wait a little while as the Internet music economy picked up steam.

“I can’t speak directly to Matchmine’s experience, but there are companies in the recommendation space that over the last couple of years raised tens and tens of millions of dollars, and we did not,” Lucchese says. “Until recently, we were five people. For a company like Matchmine, part of the problem could be that they were at full ramp-up a little bit too early for the market.” But in the last couple of months the market has really increased, Lucchese adds: “We have a very long pipeline right now of deals where people are saying ‘We need to get this out by the first quarter.’”

So what exactly is this technology that The Echo Nest’s clients are in such a rush to implement?

The company’s story starts at the MIT Media Lab, where its co-founders and co-CTOs, Tristan Jehan and Brian Whitman, met as graduate students in Professor Barry Vercoe’s “Music, Mind, and Machine” group. Jehan was developing software that automatically picked out the tempo, rhythm, and other parameters in songs. Whitman was applying text retrieval techniques to information about music—and he says he was “pretty pissed off” about the state of the art in Web-based music recommendation services.

“Amazon and almost every large online store uses sales and clickstream data to do things like saying, ‘Okay, Jen bought these four records, and you bought three of them, and you don’t know Jen, but you should probably buy that fourth one as well,’” Whitman says. That’s collaborative filtering—and it was largely invented by Media Lab professor Pattie Maes (an Xconomist) and her students in the mid-1990s (her recommendation company Firefly Network was bought by Microsoft in 1998 for $40 million).

But while collaborative filtering “sounds like a great idea,” says Whitman, “when you apply it to music, stuff just gets lost.” Music with lower sales volume, and therefore less clickstream or sales data, will never show up in a collaborative-filtering-based search. So while collaborative filtering makes tracks and albums from popular bands even more popular, it marginalizes newer or edgier groups (a point I also made in my first profile of Matchmine).

Whitman was by no means the first to find collaborative filtering unsatisfying. Back in 2000, a composer and record producer named Tim Westergren had started the Music Genome Project, an effort by scores of trained musicologists to rate songs according to nearly 400 attributes such as melody, harmony, instrumentation, and rhythm. That led to the creation of Pandora, a free streaming music service that uses the attribute scores to help users discover new music by comparing it to the tunes they already like.

Pandora is brilliant, in my personal opinion. [Editor's note: I'd have gone with "wicked awesome," but the point's a valid one nonetheless.] It offers free, 24/7 access to a very large catalog of music. The Pandora iPhone app, which makes all of that music available to mobile-device owners on the go, has brought the company hundreds of thousands of … Next Page »

Comments (2) | Reprints | Share:          

After publishing my report last night about the death of Needham, MA-based Matchmine, I wrote to CEO Mike Troiano, whom I’ve interviewed a couple of times over the past year, seeking his comments about the sudden shutdown, which he announced yesterday in the company’s blog.

Troiano wrote back this morning. He explains that Matchmine—which had built an online recommendation system based on preferences actively expressed by users, rather than on the preferences of people like them, as in the older collaborative filtering systems used on Amazon and other e-retail sites—was a direct casualty of the ongoing financial crisis on Wall Street. Troiano says he felt that after nearly two years in business, the company had finally found a formula that worked, and was about to reduce its burn rate drastically to cope with economic conditions. But the startup’s main backers, the Foxborough, MA-based Kraft Group, were forced to shut down the company to improve the group’s cash flow amidst the economic turmoil. The Krafts “made a tough business decision, and I understand it,” Troiano says.

Troiano also corrects an impression left by one sentence in his blog post yesterday in which he wrote, “It’s one thing to be failed, quite another to have been deceived.” Xconomy, TechCrunch, and others interpreted that as a hint that Matchmine had been misled by the Krafts about the status of the company’s financial backing. But that sentence “was not about the Krafts at all,” Troiano says—it was an attempt to convey to the company’s partners that the discussions going on as late as last week about prospective business deals were in good faith, since the company’s unraveling over the weekend was entirely unexpected.

Here’s Troiano’s full message:

“Hey, Wade—Saw your piece. It was a bit harsh, but more accurate and fair than most.

“I published a comment in the matchmine blog after I read the article to clarify that that sentence (‘It’s one thing to be failed, quite another to have been deceived’) was not about the Krafts at all. I was concerned that the partners we were engaged with—and the prospective partners we were still pitching as late as Thursday—might feel deceived by us. We may have failed them, in good faith, but that’s different than lying through our teeth knowing we were going to shut the company down on Monday. Hence the failure vs. deception distinction.

“As for the Krafts, they made a tough business decision, and I understand it. I still have nothing but respect for them, as business people and as human beings. In the end it really wasn’t about matchmine, it was about improving cash flow across their interests in the face of an unprecedented financial market crisis. Matchmine was just collateral damage.

“That’s the bitter pill here from our perspective. We did struggle for a while to find a formula that worked, but anyone whose tried to build a business on the back of a capability knows that’s the way it goes sometimes. Along the way we probably spent too much money, and there’s learning in that for me and my team. But in the end—just as we’d found the button, and were about to take the burn rate down just like everyone else—it ended over the course of a weekend.

“It’s going to take some time to accept that, for me, and my people.

Mike.”

Comments (1) | Reprints | Share:          

[10:00 a.m. 10/28/08: Update and clarification included below.]

In an entry cross-posted on the company blog and his personal blog this afternoon, Mike Troiano, CEO of Needham, MA-based recommendation technology startup Matchmine, said the company is shutting down, effective immediately. The failure of the company, which raised $10 million in venture financing just last year, throws at least 40 software engineers, product managers, and business development executives out of work.

The Matchmine website is no longer available, and redirects visitors to the blog, which is headed by Troiano’s message, entitled simply “Fin.” Troiano says Matchmine employees were informed of the shutdown this morning. While the company’s doors are already closed, “a few of us will stick around to try and support our partners through a transition, and notify others affected by the closing,” the post says. The company’s partners should expect calls soon, Troiano said.

Matchmine, as I explained in a profile last September, was in the midst of rolling out a system that automatically recommends music, movies, and other types of media based on the individual tastes of users, as encoded in personalized mathematical representations known as “MatchKeys.” Funded mainly by the Kraft Group, which also owns the New England Patriots, Matchmine had established partnerships with several Web-based media companies, including Fuzz, Filmcrave, Peerflix, Odeo, Blogdigger, Blogged, and MediaMelon; the companies were incorporating the MatchKey technology into their own sites’ product search and recommendation systems.

Matchmine’s path was not untroubled. In an April story following up on my original profile, I recounted how the company was forced to scrap the original system it developed for building MatchKeys as too complex and forbidding for most consumers. After the company’s public debut at the Demo conference in September 2007, six months elapsed before the company got its recommendation service working on the first two partner sites. In that time, “We made a bunch of headway and uncovered a bunch of opportunities as well as a bunch of problems with the product side of things,” Troiano told me at the time.

And while the company continued to sign up new retail partners over the summer—including ffwd.com, Pixsy, Podcast.com, Videocast.com, Muzic.com, and MP34U.com—it never scored the kind of big customer win (on the scale of a company like eMusic, say) that would have put it on the map in the digital media retailing sector.

In his blog post, Troiano indicates that while the company had been hit hard by the recent financial turmoil, its sudden shutdown was completely unanticipated. Troiano hints, without explaining, that he had been misled in some way about the startup’s financial situation. [see clarification below]

“The nature of our financing meant that the financial market crisis overtook us more abruptly than most,” Troiano writes. “To my team and their families, our vendors, network partners and prospects, I can only say that I am deeply sorry for the way this comes to a close. And I don’t mean ‘press release sorry;’ I mean really, personally sorry. Please know that I could not have imagined this last Thursday, let alone earlier. It is one thing to be failed, quite another to have been deceived.”

[Update and clarification, 10/28/08: In a response today to my inquiry about that last sentence, Troiano says he was not trying to imply anything about the company's financial backers. "That sentence...was not about the Krafts at all," Troiano wrote to me this morning (see my followup story). "I was concerned that the partners we were engaged with---and the prospective partners we were still pitching as late as Thursday---might feel deceived by us. We may have failed them, in good faith, but that’s different than lying through our teeth knowing we were going to shut the company down on Monday. Hence the failure vs. deception distinction."]

The bulk of Troiano’s post, however, is devoted to a list of Matchmine’s employees, whom other companies “smart enough to be greedy while so many are scared” would do well to snatch up, he argues. These include Trent Adams, the company’s founder and “Chief Innovator,” who first conceived the recommendation technology behind Matchmine’s system while working directly for the Kraft Group.

It’s possible that Matchmine’s recommendation technology will again see the light of day: Troiano says in the post that the company’s intellectual property is now for sale. “We’re open to selling the IP and patent portfolio, and just as willing and able to protect it,” he writes.

Comments (1) | Reprints | Share:          

Seattle-based Evri has announced its beta site is open and ready for user feedback. Evri, which is backed by Vulcan Capital, wants to make the Web better for browsing by understanding connections between people, products, places, and things—and providing pages and widgets to recommend relevant content.

Comments | Reprints | Share:          

Matchmine—the Foxborough, MA-based startup whose recommendation engine helps consumers find songs, videos, and other content they’re likely to enjoy at various content-oriented websites, and which we profiled in September 2007 and again in April 2008—announced today that four new destination sites have signed up for its system. They are Podcast.com and Videocast.com, both owned by Treedia Labs, and Muzic.com and MP34U.com, both owned by MP3 Lunatics. Earlier this month, Matchmine also announced partnerships with personalized channel-surfing site ffwd.com, and Pixsy, a maker of private label video and image search engines.

Comments | Reprints | Share:          

When we first wrote about Needham, MA-based Matchmine last September, the Kraft Group spinoff had just introduced its MatchKey “portable preferences” system, which is designed to put an end to the fragmented state of recommendation systems on the Web. Right now, every online retailer, video rental site, or music site you visit—be it Amazon, Netflix, or iTunes—has a separate system for recording what types of products you like and recommending new ones based on your preferences. Matchmine’s concept is to consolidate your preferences into a single mathematical representation, your MatchKey, that would follow you across many sites and many types of products and media, sparing you the hassle of educating each site separately—and sparing smaller online businesses who join the MatchKey network the need to create in-house recommendation systems at all.

The idea sounded great, but in practice, there have been some bumps along the way. Matchmine announced at the Demo conference in San Diego last September that its first two commercial partners would be Fuzz, a guide to the local music scene in San Francisco and Seattle, and movie rating site FilmCrave. But it took another six months to get the MatchKey service working on their sites (it went live on both sites on March 31).

In that time, “We made a bunch of headway and uncovered a bunch of opportunities as well as a bunch of problems with the product side of things,” says Matchmine CEO Mike Troiano. For one thing, the company found that the process it had engineered for creating a MatchKey—which involved going to a central website, answering a long series of questions about one’s media preferences, and then downloading some software—created much too forbidding a barrier for most potential users. Also, it was taking too long for Matchmine to sign up new media-site partners, process their catalogs, and integrate the MatchKey scheme with their existing account-management systems.

This week Matchmine seems to be putting its long winter behind it and taking some big steps toward spreading its technology. Yesterday it named four new media partners—music, podcast, and video aggregation site Odeo, blog search engine Blogdigger (an Odeo subsidiary), blog directory Blogged, and high-definition video site MediaMelon—and promises to have the MatchKey system up and running on those sites “in a matter of weeks, not months,” in Troiano’s words.

The company, which emerged from the same part of the Kraft Group responsible for the New England Patriots’ website and multimedia offerings, also detailed the improvements it’s made to its platform to make it easier to use—both for consumers who want to sign up for MatchKeys and for new commercial partners who want to get the system working with their sites. “It’s night and day, in terms of how easy it is now to get started,” says Troiano.

For one thing, Matchmine has done away with the downloadable software. Now users’ MatchKeys live entirely online, and are accessed behind the scenes on each participating website to guide the local recommendation engine. If you go to Fuzz or Filmcrave and you don’t yet have a MatchKey, you can easily create one that applies just to that site by answering just a few questions about your music or movie preferences. Later, you can decide whether to “promote” your local MatchKey so that it can keep track of your preferences across many sites. (The company has set up a new website that explains to potential users how the consumer-facing side of the system works.)

Though MatchKeys don’t yet work with top online brands like Netflix, Amazon, or Last.fm, Troiano says the company will soon roll out a feature that allows users to train their MatchKeys quickly by, in essence, sucking up the information in the preferences and ratings they’ve already expressed at such sites. (Many major consumer sites have public application programming interfaces or APIs that allow outside programmers to extract and re-use this kind of information.) “That would enable me, for example, to make sure that the preferences I’ve expressed at Netflix are … Next Page »

Comments | Reprints | Share:          

Remember when computer scientists promised back in the 1990s that we’d all have artificially intelligent “personal agents” running cyberspace errands for us, booking travel, filtering the news, and finding things we like? A new widget for personal computers called a “MatchKey” doesn’t quite qualify as intelligent, but it does take a new approach to the third problem—sifting through the Internet’s vast catalog of content and picking out items that match your preferences. It has the potential to create far more accurate music, movie, and product recommendations than the ones Web users receive today, and it has emerged from the unlikeliest of places: the company that owns the New England Patriots football team.

At the DemoFall conference in San Diego yesterday, Needham-based startup Matchmine unveiled what it calls a “media discovery platform” built around MatchKeys—mathematical representations of individual users’ preferences in areas such as music, movies, and other online media. Once Web surfers have signed up for their own MatchKeys, the keys will monitor their media-consuming habits and continuously update their preference scores. Whenever users visit Matchmine partner sites such as Fuzz, Filmcrave, and Peerflix, software at the sites will tap into their MatchKeys and present content fitting their profiles.

“The big idea is that we want to make online media personalization portable,” says Mike Troiano, CEO of Matchmine, which also announced Monday that it has collected $10 million in Series A venture funding from The Kraft Group, the Foxborough, MA, company that owns the Patriots as well as paper manufacturing and real estate development interests.

The Patriots were among the first NFL teams to develop an extensive website with multimedia offerings such as a streaming radio station. Troiano says Matchmine grew out of a conversation in 2005 between Kraft Group president Jonathan Kraft and Trent Adams, the group’s “chief innovator,” about technologies that would help people find the Patriots’ content online.

“Trent realized that this is a problem for every user and every publisher: If you’re a user, how do you find interesting stuff, and if you’re a publisher, how do you connect with people who will be interested in your stuff,” Troiano explains. “Jon and Trent thought that if we could crack that nut, it would be a hugely profitable business. And within a couple of weeks, Trent had this epiphany about how to create a statistical representation of users’ tastes. That’s the core that became the MatchKey.”

Today, websites such as Amazon generate product recommendations using collaborative filtering software, which works on the principle of shared taste: “You bought this Annie Lennox CD, and a lot of people who like Annie Lennox also like Melissa Etheridge, so you’ll probably like this Melissa Etheridge CD.” But the basic algorithms behind collaborative filtering are nearly a decade old, and have many limitations, including the fact that they reinforce popular choices while marginalizing unusual or lesser-known content.

The information in a MatchKey, by contrast, isn’t about specific products or other people’s preferences. It’s essentially a listing of a user’s scores on 200 axes across different media types, including movies, online video, music, and blogs. Each media type has 50 to 75 axes; for movies, for example, an individual might have a low score on an axis such as “likes science fiction” and a high score on axes such as “likes mobster movies,” “likes violent movies,” and “likes movies with lots of celebrities.”

The other half of the equation is Matchmine’s “universal content catalog,” which ranks the items in its partner companies’ databases along the same axes found in the MatchKeys. When a MatchKey owner visits a Matchmine partner site, the site grabs the ID for the user’s MatchKey and sends it to Matchmine’s media discovery server, which reaches into the universal content catalog and sends back content tailored to that key. The individual mentioned above, for example, might get a recommendation for the movie Goodfellas.

To give users a sense of connection to their MatchKeys, Matchmine has also developed a graphical representation of the key, a spiky crystal sphere that looks like it was swiped from Superman’s Fortress of Solitude. “We want people to connect with their key on an emotional level and give them a sense of ownership,” says Troiano. The lengths and configurations of the spikes on a MatchKey are unique to each user, and while their meaning isn’t exactly clear, the company is working on visualization tools that will make the keys more meaningful to the general user. Says Troiano, “Some of our scientists can already take one look at a MatchKey and say, ‘This is a big sci-fi fan.’ But you can imagine tools like a 3-D flyover of your music zone with rollovers showing what each crystal means.”

Such baubles might be a nice distraction. But if there’s real value in a portable preference engine like Matchmine’s system, it’s that it could help consumers move to the next stage in media personalization—unifying their preferences in one place rather than leaving them scattered across dozens of separate e-retailing sites, where the last few albums or movies a user bought might be all the retailer knows about them.

Just as important, MatchKeys are designed to get smarter over time. “As you consume content in any Matchmine-enabled application, your key will evolve based on the preferences you express implicitly and explicitly,” says Troiano. “What that means is that the time you invest in any one discovery experience will benefit you in all the others.”

At the moment, there aren’t that many opportunities for discovery experiences through Matchmine, which launched with only three partner sites—artist-centric music sharing site Fuzz, user-written movie review site Filmcrave, and DVD-trading site Peerflix. But Troiano says Matchmine will be adding more partners “at a feverish pace” over the next few months. “What we’re finding as we talk to these folks is that they know they need to invest in the development of some kind of attribute-based recommendation system, but they just don’t have the resources or manpower, whereas we have a large organization dedicated to doing nothing but that,” Troiano says. “More often than not, they are eager to have us come in and score their content catalog and begin to make recommendations, because they know they have to do it.”

Comments (1) | Reprints | Share: