February 8, 2012
I want to preface this post with the following: this is in no way meant to be vindictive, and I fully respect the work and time gone into develop each of these products. I don’t believe any of these products to be “useless”, as some readers have suggested, and this blog post is more of an analysis of the most recent batch of travel startups and why none of them has really captured my imagination. They all have users, so I realize they are useful, but this is the way I see it.
UPDATE: Both Gtrot and Trippy have pivoted since this article was written. Gtrot is now more “local discovery” and Trippy is essentially a “Pinterest for Travel.”
The original article:
I don’t use the vast majority of new travel startups. If there is anyone who should, it would be me: I’m a bay area techie who has traveled to 65 countries, has dual citizenship, and probably been on close to 200 flights. But I don’t, and here is why:
In my opinion, they are marginal improvements on current solutions, or attempts at being “social” without doing any hardcore hacking. The hype surrounding them is overblown and often based on how pretty something is, how “social” it is, or which VC or angel invested.
Here are the reasons I have not become an enthusiastic early adopter:
1) They are based around a process that isn’t that difficult to begin with
Many startups are focusing on relatively small problems, but approaching them way too aggressively . Where the problem might require a feature, or a widget, we get full blown social networks. They use a sledge hammer where a chisel would be more appropriate. I see Gogobot, Trippy and Dopplr as the biggest examples of this.
Gogobot allows you to “see where friends have traveled” and “get tips from travelers like you.” The way to solicit advice on Gogobot is to post a trip. Personally, I guess I don’t see why posting a facebook status is all that different. I can see where friends have traveled through their facebook albums, and I can ask them advice via facebook.
Just now on the Gogobot page, it has notified me that Felipe (who I don’t know) has planned a trip to Rio. Why am I better equipped because of Gogobot to give this guy advice than say, his facebook friend? It seems like Gogobot just formalized the process of making a “Trip” and giving advice on it, and created an entire social network around it.
Trippy has the same problem. I sign on, and it shows me that a few of my friends are planning trips to Colombia etc. But my friends (or myself even) don’t travel often enough that I have incentive to sign in on a regular basis.
The problem of knowing my friends upcoming trips should not be solved by an entire separate social network: it is not important enough. Social networks for travel like Couchsurfing/Tripping work because they aren’t just about the trip: they are about hosting other travelers and being part of a movement and community. Trippy and Gogobot don’t have that quality.
Dopplr allows you to “share your personal and business travel plans with the people you trust”, basically allowing you to find out when friends are in the same area. I can see how this may be useful, but to me, sending an email to my friends works just fine.
Dopplr, Trippy and Gogobot are just a few of many companies out there that, for me, offer very little that you couldn’t currently get out of facebook, email etc., and the pain they are solving isn’t big enough to add yet another social network or service to my life. Perhaps for the wandering businessman or gap year traveler, these services prove invaluable, and they definitely have potential.
2) “Social Travel” is currently broken
These companies are tackling an admirable problem, but their algorithms need to be better.
I just used Gtrot for San Francisco, and it showed me, among other tourist destinations, that my friends had gone to The Fillmore, Tipsy Pig, Candlestick Park and an advertisement for two tickets to the Stanford Cardinal. I don’t really like live music that much (Fillmore is for concerts), I hate the Tipsy Pig because it is way too crowded, I don’t follow the 49ers or professional football (they play at Candlestick) and I went to UC Berkeley! If I didn’t know all this already and was visiting, Gtrot would have implicitly recommended that I go a bunch of places that I would have hated. . . .
Foursquare Explore is even worse: telling me that Wrigley field is where the most people checked in in Chicago is useless . . . I could have looked that up on wikitravel.
I don’t fully agree with this guy, that the social travel model will NEVER work. But he makes a good point when he talks about how, even among close friends, travel preferences can differ drastically.
Eventually, if Gtrot can figure out an algorithm that uses the social graph well then I would be interested, because I like their interface and the concept is good. My 6 closest friends wouldn’t be caught dead in the Tipsy Pig, don’t really follow much sports, went to Berkeley, and also don’t really go to concerts that much. If a company was able to figure that out and recommend locations based on those friends, that would be useful. If they could figure out I’m a Cal Bear and not recommend things having to do with our rival school that would be an improvement.
This is a big data problem, NOT a social problem. Someone needs to be pursuing machine learning algorithms to learn from all the social data out there. Facebook has made it all available, but no startup has really analyzed that data past “your friends like this > you may like this.” Analyze which friends, analyze how often we post on each others walls, our mutual interests, how much space on my best friends wall I take up compared to other friends. These companies could already be doing this, but as of now it isn’t apparent.
3) They aren’t different enough
Hipmunk. Don’t get me wrong, the interface is beautiful. However, I don’t use it.
When we were developing Mozio’s interface, we considered a Hipmunk-esque style approach. In fact, we are still considering a Hipmunk-esque approach. We did a fair amount of introspection, asking ourselves, why don’t we use Hipmunk? I still automatically go to Kayak for flights. I realized that the main thing Hipmunk has going for it is the ability to visualize the times of the flights.
This is pretty cool, I won’t deny, and it looks gorgeous. However, I realized, I don’t NEED it. It’s nice, but more along the lines of “oh that’s cool”, not “oh wow I have always felt I really needed this”. I feel like this is almost heresy to say, since Hipmunk is a YC favorite and Ashton Kutcher invested. And the founders are from Reddit, I get it. But I am not drinking the coolaid.
I realized I’m perfectly fine using the sliders on the left of the Kayak results page. Being able to visualize that my flight lands at 6pm is no better to me than just adjusting the sliders so that no flight is displayed that lands after 6pm.
Hipmunk is a useful product . . . I want to make that clear. I like the design, but it is just maybe only slightly more useful than what is already out there, and not solving a big enough customer pain for me to remember to type hipmunk.com into the address bar instead of going straight to Kayak. (I note though that they have started integrating trains, which I fully support and is something we at Mozio are doing as well.)
If I’m lucky enough to have anyone from these companies see this blog post, maybe there will be a spirited discussion about how to use their product. Looking forward to it.