From MIT Blackjack to E-Mail Databases: We Catch Up with the Other Micky Rosa
Lots of folks are excited about the movie 21, a semi-fictionalized account of the exploits of the MIT blackjack team in the 1990s. That’s clear not just from the film’s box-office-leading performance last weekend, but from the fact that so many people are coming out of the woodwork and identifying themselves as members of the team.
In fact, there’s a bit of a controversy brewing in the comments section of Bob’s piece last week about team member John Chang, one of the models for the Micky Rosa character in Ben Mezrich’s 2002 book Bringing Down the House, upon which the movie is based. Bill Kaplan, who co-founded, trained, and financed the blackjack team, commented that, “While I know John likes to say he’s Mickey Rosa…the fact is he’s a dead ringer for the character Choi.” (Note: the book spelled Rosa’s first name “Micky” but everyone routinely adds an “e.”) And later Kaplan e-mailed Bob the following note:
“The Mickey Rosa character as developed in the book and the movie is fictitious. But as he represents the founder and leader of the team who made everything happen (e.g. supplied/raised the capital; determined the original playing strategy; structured the compensation terms and incentives for investors, players, and management; put the business processes, training rules, checkout procedures, player tracking and supervision, etc. in place; gone out as a player on top), that would be me.”
We aren’t about to take sides on this one. But it’s obvious that the Rosa character (played by Kevin Spacey in the movie) is likely even more of a fictionalized amalgam than has previously been reported. And what’s also becoming clear is just how many of the people who did time on the blackjack team have gone on to become prominent members of the Boston technology and innovation scene. Kaplan is a prime example of that, as I learned in an interview with him this week.
Kaplan is a 1977 economics graduate of Harvard College (the same class that included Steve Ballmer, and that would have included Bill Gates, had he not dropped out). Among the many feathers in his cap are a former world ranking in professional squash and a 1980 MBA from Harvard Business School—he had to fight to get his initial admission to HBS reinstated after the school found out about his history as a card counter.
Kaplan says he co-founded the blackjack team in 1980 with MIT student J.P. Massar (who went on to be a well-known LISP programmer, an employee at Danny Hillis’s pioneering computer company Thinking Machines, and a professional-level poker player). Kaplan ran the club as a limited partnership bankrolled by outside investors; at its peak in the early 1990s, it “employed” as many 80 players, Kaplan says. He departed around 1993—before most of the events depicted in Bringing Down the House—to focus on his Boston-based real estate investment company, Linden Properties, which he had started in 1980 and which has acquired and developed over $100 million in commercial real estate, most of it within a two-hour drive of the city.
And for his latest act, Kaplan has remade himself again, this time as an Internet entrepreneur. The problem Kaplan and his Newton, MA-based company, FreshAddress, are trying to solve is one endemic to the net, where people are even more transient than in the physical world. It’s the fact that people switch e-mail addresses every three years or so, on average. While that may not seem like such a big deal from a consumer’s point of view, it causes fits for any organization that depends on e-mail to communicate with its customers, donors, or alumni. It means that if a company assembles a database of 3 million e-mail addresses, after a year only 2 million of them will be usable. And that’s something companies are willing to pay to prevent.
When Kaplan and business partner Austin Bliss started FreshAddress in 1999, however, they were thinking of it as a consumer service. “Austin came up with this idea of an email change-of-address company, providing the missing link between someone’s old address and their preferred address for people who have lost touch with an old associate or classmate,” says Kaplan. The company won a patent on the idea of creating a database of “change pairs”—matched electronic identifiers such as e-mail addresses—and created a consumer website where users could supply the company with such change pairs by entering their old and new addresses.
“But my concern was that you can’t make money off that,” Kaplan recounts. “Consumers weren’t willing to pay anything. Everybody thought that just bringing eyeballs to your website was so great—and a lot of those companies went under. But what’s more interesting is that companies who are losing touch with their customers should be willing to pay money to reconnect with them.”
So FreshAddress spent a couple of years rebuilding its technology as a corporate service, which it started selling in 2003. FreshAddress’s clients start by sending in lists of inactive or bouncing e-mail addresses from their customer databases. (The company only works with customers who can show they have a legitimate relationship with the people on these lists, and that the original addresses were volunteered or obtained through an opt-in process. In other words, no spammers.) Sometimes e-mail addresses are bouncing simply because they contain errors: somebody typed “firstname.lastname@example.org” into an online form instead of “email@example.com.” By applying automated “hygiene” algorithms, FreshAddress is able to correct half to two-thirds of those problems, Kaplan says.
If it’s not a hygiene problem, FreshAddress looks for matches in its database of change pairs. That database is drawn from the company’s own consumer-facing website and from databases that the company licenses from hundreds of other websites where consumers have opted to provide their e-mail addresses.
“Say you gave the address firstname.lastname@example.org to Dell at some point,” Kaplan explained to me. “Now it’s bouncing. I do a lookup and I see that now you’re at email@example.com, or whatever. We send out an e-mail on behalf of our client to confirm that e-mail is deliverable to the address we’ve matched, and we give people a chance to opt out if they don’t want to reconnect. We then deliver the updated e-mail addresses to our clients on a pay-for-performance model. We only charge for the guaranteed, delivering e-mail addresses.”
Lately, the company has added another service: finding e-mail addresses in cases where a client only has a customer’s postal address. “A general rule of thumb is that most companies have e-mail addresses for only about 25 percent of their postal database; about 75 percent are missing,” says Kaplan. “And on the e-mail side, about 30 percent of one’s file will start bouncing through attrition over the course of a year. Those are pretty big numbers, and it’s a huge cost to a company.”
Indeed, FreshAddress’s service is proving so popular that revenues at the self-funded, 20-employee company grew by 50 percent between 2006 and 2007, Kaplan says, although he declines to cite specific revenue figures. And the phenomenon that makes FreshAddress’s business possible isn’t going away anytime soon. “There are a few people who have permanent e-mail addresses”—for example through their college’s alumni organization—“but the fact is that people are switching schools, switching jobs, and switching e-mail providers al the time, looking for a better deal,” says Kaplan. “They’re switching from AOL to Gmail, or from Comcast to Verizon, or they have a Prodigy address which becomes SBC which becomes AT&T.”
To complicate matters further, most active Internet users have at least three separate e-mail addresses, which they give out for work purposes, for personal matters, and for newsletters and commercial offers. Somebody needs to sort it all out. And that’s the humble task that will keep FreshAddress occupied for a few years—or at least until someone figures out how to supply consumers with secure, universal, persistent identities on the Internet (and there are plenty of people working on that, including the OpenID Foundation).
Kaplan says FreshAddress had only one major competitor—a New York startup called Return Path that, he says, spent a lot of its venture funding “running full page ads in all of the direct-marketing magazines, which was honestly very helpful for us in terms of educating the market.” But after FreshAddress won its patent, Kaplan says, ReturnPath moved mainly into e-mail marketing consulting, and is now one of the companies that licenses its address data to FreshAddress.
“We just completely blew them out of the water,” Kaplan says, showing a bit of the competitive spirit that was undoubtedly helpful during his squash- and blackjack-playing days. Indeed, judging from his past, Kaplan no doubt thinks FreshAddress has better than house odds of eventually conquering the e-mail attrition problem.