Hipmunk, Conceived by David Pogue’s Teenage Co-Author, Embarks On Mission to Make Travel Search Easier

8/18/10Follow @wroush

This is the fourth in a series of profiles of companies emerging this summer from Mountain View, CA-based startup incubator Y Combinator.

If anyone was ever predestined to be the co-founder of a Y Combinator-backed Web startup, it’s Adam Goldstein.

He’s got the academic pedigree, having gotten his bachelor’s degree from MIT in electrical engineering and computer science this spring. And he’s got the connections. He has known Y Combinator founders Paul Graham and Jessica Livingston since he was 16 years old. He’s a friend of O’Reilly Media founder Tim O’Reilly and is the author of two O’Reilly books, one co-authored with New York Times tech columnist David Pogue. And he’s got some entrepreneurial experience. Together with Wired editor Chris Anderson, he co-founded BookTour, a site that assists authors with book promotions. Goldstein has even got some of his own seed capital. He tells me that he’s been saving money since he was 16 so that “were I ever to decide to start a company on my own, I could afford to not draw a salary for several years.”

Well, hopefully it won’t be several years before the 22-year-old Goldstein can start drawing a salary at Hipmunk, the Y Combinator-backed company that he launched yesterday together with Reddit co-founder Steve Huffman. The company is attacking the user-interface side of online travel planning, attempting to bring some rationality and convenience to notoriously vexing tasks like booking flights and finding hotels and rental cars.

Currently, Hipmunk’s website focuses on flight searches. The user starts out in the usual way, by specifying origin and destination cities and dates for an upcoming plane trip. But rather than spitting back page after page of results in no discernible order, the way many flight search engines do, Hipmunk will filter out the bad options and sort the remainder on an intuitive “agony” scale that takes into account price, duration, and number of stops. “We’re going to sort the results sensibly, filter out the flights that you’d be crazy to take, and present things in such a way that it’s clear what the tradeoffs are,” says Goldstein.

Hipmunk flight search resultsThat’s a vision Goldstein says he has been incubating for more than a year, since his days as travel planner for the MIT debate team. “I had spent the past four years booking travel for the team and it’s incredibly painful,” he says. “I spent hours looking through Kayak and Orbitz and Expedia trying to find the right flight. It doesn’t have to be that way, but their interfaces really aren’t made for anything else. It’s just a huge database dump.”

Goldstein wants Hipmunk to become known as the intelligent travel search engine—the one that knows that if there’s a non-stop flight from San Francisco to New York that costs just $1 more than the one-stop flight, it probably shouldn’t be buried on page 5 of the search results.

“I think other companies have really dropped the ball on flight search,” Goldstein summarizes. “When Kayak came out, a lot of smart people thought it was the greatest thing to happen to flight search in years, and I think they were right. But looking back at it, I feel like the fundamental experience for most users didn’t change very much. You still have to short through the redundancies, the code-share flights, and all that stuff. People will use a better search engine, if it exists.”

That’s the basic story behind Hipmunk, which unveiled the flight search engine to the public on Tuesday and is now collecting user feedback about bugs and requests for added features. I’ll get back to that technology in a moment. But to me, there are two very interesting threads to the company’s story: first, Goldstein’s path to Y Combinator, where he and Huffman are part of a group of 36 companies that will soon complete the startup incubator’s summer term; and second, the challenges Hipmunk faces as a newcomer in the travel search business, a crowded field with some well-established competitors.

I met up with Goldstein at a SoMa coffeehouse yesterday, and if I repeat a bit of the life story he shared with me, you’ll understand why I say he was predestined to be a startup founder. A programmer from his early teen years, Goldstein was an admirer of Pogue’s Missing Manual series for O’Reilly. He met Pogue at a book signing when he was 14, and the columnist hooked him up with O’Reilly, which led to an authoring assignment—AppleScript: The Missing Manual—and to some editing work with Pogue. When Goldstein was 16, he and Pogue co-authored Switching to the Mac, a guide to the Mac OS X operating system for reformed Windows users.

Goldstein spent the summer of 2005, between his junior and senior years in high school, working at O’Reilly Media in Sebastopol, CA, where Tim O’Reilly invited him to the company’s exclusive Foo Camp hacker event. It was the same summer that Graham and Livingston were launching Y Combinator—a combination startup school and venture investing operation—and at Foo Camp Goldstein met both Graham and Huffman, who was starting social bookmarking site Reddit as part of the very first class of Y Combinator companies.

During his freshman year at MIT, says Goldstein, “Out of the blue I got an e-mail from Paul Graham, who said ‘Chris Anderson is looking to start a company, you should talk to him.’” Anderson, of course, is the Wired editor whose bestseller The Long Tail appeared in 2006. Anderson wanted to start a company that would help authors find audiences. “As someone who had been an author, I thought that was a good idea,” Goldstein says. So he co-founded BookTour with Anderson, and dived in on the project full time during the summer before his sophomore year of college.

At that time, BookTour was borrowing office space from Reddit, which Condé Nast, Anderson’s employer, had just acquired. It was one of the first dramatic exits for any Y Combinator-backed company. Goldstein ended up not just sharing an office with Reddit, but rooming with Huffman.

Fast-forward to early 2010. Huffman had left Condé Nast in 2009, and Goldstein, who was about to finish his undergraduate work at MIT, had given notice at BookTour, which “hadn’t been getting a huge amount of traction,” in his words. “One day I called up Steve and said ‘I want to work on a startup, what are you doing this summer?’” he says.

Huffman liked Goldstein’s pitch for a business built around a more intelligent travel search interface. In March, the two attended “Demo Day,” the climactic investor-pitch session at the end of Y Combinator’s winter 2009-2010 term, then pitched their own idea to Graham. Graham went for it, and invited Huffman and Goldstein to join the summer 2010 session. That made Huffman a repeat Y Combinator startup founder—not the first, evidently, but one of only a handful, according to Goldstein.

Given Huffman’s success with Reddit, Hipmunk “could have gotten funding anywhere,” says Goldstein. But he says there were strong reasons to work with Y Combinator, despite the fact that the capital startups receive there is relatively expensive. (The average Y Combinator startup gives up 6 or 7 percent of its founding equity in return for a stipend amounting to about $11,000 plus $3,000 per founder.) “Whether we took money from them or not, Paul Graham is a fan of Steve’s and mine, and we knew he would have helped us out either away, and it just seemed wrong to take advantage of that,” Goldstein says. “It was better to get on board with them. And these days there is a great deal of prestige that comes from being in Y Combinator.”

Indeed: “super-angel” Ron Conway, at a recent Y Combinator-sponsored conference on angel investing, called Y Combinator “the MIT, the Harvard, or the Stanford of incubators.” Conway, you may not be surprised to learn, is also an investor in Hipmunk, through his firm SV Angel.

That’s the back story at Hipmunk, surely one of the most highly pedigreed companies ever to emerge from Y Combinator. But what about its future? Goldstein is probably right that there’s room for improvement in the way consumers book flights and make other travel arrangements online. But having profiled Boston-area companies like ITA Software and TripAdvisor, both powerhouses in the travel search arena, I happen to know that the amount of brainpower already focused on this exact problem is spectacular. ITA, in particular, is full of MIT-trained brainiacs who eat and breathe flight search algorithms, and have made their company the dominant provider of flight information to other companies like Orbitz. (Which is probably why Google is in the process of acquiring the outfit for $700 million.) What makes Goldstein think that he and Huffman have the key to the next generation of travel search?

“It’s all about the user experience,” Goldstein says. “The back end problems, you’re right, are big, big problems, and there are people a lot smarter than me working on them. I think the opportunity is to make the presentation of the information more easily understandable and more useful. That’s why we decided to start there.”

Goldstein argues that the competition is moribund. Orbitz, Kayak, Expedia, and other companies have been offering the same graphical views of flight options, often with ITA’s software under the hood, for years. Last year, TripAdvisor added a new flight meta-search service to its traditional hotel and restaurant guide—and in an odd coincidence, one of the people who built it, Brian Krausz, is a classmate of Goldstein and Huffman’s at Y Combinator this summer, working on his own startup, GazeHawk. But while Goldstein says that while he appreciates “the thought they seem to be putting into flights,” he doesn’t see the TripAdvisor offering as a qualitative improvement over other flight search options.

Goldstein and Huffman—who have split up the coding work on Hipmunk about 40/60, Goldstein says—have clearly put a lot of work into their own interface. Flight data, which currently comes exclusively from Orbitz (which means it actually comes from ITA), shows up extremely quickly, and is arrayed using time bars that make it easy to visualize flight departure and arrival times. When there’s a cluster of similar flights available, the interface shows only the one with an optimal combination of price, duration, and layovers, and hides the others, though they can be easily revealed. A row of buttons along the top of the screen lets the user instantly re-sort flights by price, duration, number of stops, departure time, or arrival time. Draggable sliders let the user adjust the acceptable departure or arrival times, and there’s even a tab system that allows users to keep multiple searches open in different tabs.

ITA Software's flight search matrixIt’s an impressive effort that obviously involved a lot of Python and JavaScript coding. But overall, Hipmunk’s user interface is very similar to a search results matrix developed by ITA Software (see graphic at left). ITA hasn’t ever seriously marketed this interface, and as far as I can tell, it’s a secret to all but the geekiest travelers. But the point is that Hipmunk is going up against a 500-employee company that’s about to become part of a 20,000-employee company.

Goldstein’s reply comes in two parts. First, he says, Hipmunk has thought harder than most other flight-search providers about usability. “We think [ITA's matrix] is a cool interface, but we think we have a lot of stuff that differentiates us, like the draggable sliders for the time of day, hiding the flights that you’re not interested in,” he says.

Second, he says it would be premature to judge Hipmunk based just on the product it launched Tuesday, which, if you wanted to boil it down, could be described as merely an improved interface for flight search. Which gets to an interesting paradox that afflicts almost all Y Combinator companies: Graham and the other mentors involved in the startup incubator relentlessly push YC companies to launch products early and “iterate,” that is, come out with serial improvements based on user feedback. That’s a perfectly good way for a software startup to get through its early stages, but it does put the incubatees in the position of having to plead for customers’ (and journalists’) patience while they iterate toward something weighty and worthwhile.

Goldstein insists that there’s much more to the Hipmunk vision than meets the eye. “The broadest frame here would be that we want to help people find what they are looking for,” says Goldstein. “A little more narrowly, it would be helping people understand what the tradeoffs are in the world of transportation and tourism. And then, narrowing down even more, what we have right now is a tool that helps people understand the tradeoffs in flights.”

The trick for early users—not to mention potential investors—is to look at Hipmunk’s current tool and then zoom back out to the bigger frame, imagining what might become possible if Goldstein and Huffman are given the time and resources to iterate on their vision of rationalizing the travel search business. If they can conquer flight search, after all, there’s a lot more data that needs to be organized and displayed better: hotel rooms, rental cars, restaurant reservations, and basically any commodity where people need to weigh prices, schedules, availability, and other data all at once.

Time will tell whether it’s Goldstein’s destiny to solve such problems. But he’s going to try. “Everything in my life has been leading up to working on a startup,” he says. “Sometimes people ask why I didn’t drop out of college to start a company, and I say that I never really found an idea that was good enough to drop out for. If I’d found this idea three years ago, I might have.”

Wade Roush is a contributing editor at Xconomy. Follow @wroush

By posting a comment, you agree to our terms and conditions.

  • http://danweinreb.org/blog Daniel L Weinreb

    I’d like to comment on the general concept that it’s best to get out to customers, get feedback, and iterate as quickly as possible. There are a lot of good things to be said about that approach, and they’be been said in many places, so I won’t repeat them here.

    There’s the drawback that Wade mentions. But also: “You only get one chance to make a first impression.” If someone comes and sees your newly released “0.2.1 beta”, which has fewer features and more bugs than you would like, it’s possible that she will say “This site is no good”, and never come back, even when you’ve fixed the problems.

    It seems to me that to make the “get out there early and iterate” strategy work, it’s a lot safer to get out there early with only a small number of users, chosen to be friendly enough to not bad-mouth you, but honest enough to give you uncensored feedback. During this period, you should not publicize yourself. For example, you should tell Xconomy that you’re not ready yet. You should avoid buzz and hype. Save that up for when the product is “ready for prime time”, i.e. ready for lots of users who you did not hand-pick.

    This is entirely my personal opinion about the strategy in general, not about Hipmunk in particular. It has nothing to do with (full disclosure) my being an ITA Software employee nor an Xcomomy investor.

  • http://www.roomatlas.com Roger Middleton

    Love it. A true “why hasn’t anyone done this?” moment for me. If they can expand the range of providers to be competitive with Kayak and Skyscanner I’d make this my first and only stop.

    Speaking of travel visualizations… a shameless plug for my own “Hotels.com hotels on Google Maps” site: http://www.roomatlas.com. Live prices in the markers, Street View and Wikipedia placemarks.

  • http://grumomedia.com Miguel Hernandez

    Hipmunk’s interface has improved considerably after this post.
    I created a little video showing this new and improved interface and also explaining what Hipmunk is all about.
    The video is here -> http://youtube.com/watch?v=6teBPUgz4Y8
    Regards, Miguel.