Hipmunk, the flight and hotel metasearch company, enjoyed an extraordinary blitz of West Coast tech media fanfare after its launch in August 2010.
Yet many long-time members of the travel industry remained skeptical, having seen metasearch sites — and startups generally — come and go. Some considered the site to be the pour cousin of American metasearch giant Kayak.
Kayak’s CEO, Steve Hafner, has voiced the opinion of many skeptics by variously calling Hipmunk “roadkill”, “irrelevant”, and “unremarkable“.
Hipmunk remained a darling of the tech press, thanks partly to its user experience (which emphasizes bar charts and heat maps), partly to its founders having graduated from the Y Combinator startup incubator program, and partly from the founders’ backgrounds in having helped create the internet community’s favorite socially-driven news site Reddit.
Yet the travel establishment continued to look at the startup with raised eyebrows, waiting for a hockey-stick-shaped, spiking trajectory of revenue growth — which has never appeared.
In a roughly overlapping timespan, Airbnb (a fellow Y Combinator graduate) has experienced sensational growth. Companies with comparable metasearch products like Momondo and Skyscanner have claimed much faster spurts of growth.
Reporting “insane” mobile numbers
Against this context, Hipmunk has spent the past year emphasizing mobile growth. And this strategy seems to have worked.
Strong organic growth is what CEO Adam Goldstein talks about when he talks about Hipmunk today.
Hipmunk’s mobile usage numbers are “going nuts,” he says. Its apps and mobile website together drive “close to half” of the company’s traffic. Its Apple app has seen “nine times year-on-year growth,” he adds.
This growth isn’t a lucky break, says Goldstein. It’s a benefit earned because of Hipmunk’s recent emphasis on mobile. In the past year it has overhauled its apps for Apple and Android devices. It also debuted its first mobile-optimized site.
Goldstein credits the August introduction of Tonight deals — a metasearch pull of same-day hotel offers — for amplifying the mobile growth.
The company’s hiring pattern suggests that it feels confident, three and a half years since its high profile launch [TLabs here]. In 2013, it added 10 employees, most of whom were engineers. Today it has 36 people. Goldstein says he hopes to hire 15 to 20 more this year.
Given those hiring goals, a new round of funding seems likely. The company has raised $20.2 million to date (the last round being a hefty $15 million in June 2012).
For an update on the company, Tnooz chatted by phone with Goldstein. While he didn’t reveal specific revenue and traffic numbers, he did provide some hints about the company’s track record, where its revenue is coming from, and its plans for 2014.
It was a wide-ranging discussion that also covered broader industry trends, such as how customers search on mobile differently and how the proposal of a major airline group for a “New Distribution Capability” might affect metasearch.
Some critics say Hipmunk hasn’t scaled quickly. Therefore you either built the wrong product or you’re marketing it poorly. Thoughts?
There’s no debate that we’re smaller than the OTAs [online travel agencies], Kayak, etc. But *they* were also small when they were only three-and-a-half-years old.
In terms of where we are relative to where we expected to be, we’re right on track. Travel is not a network-effect industry. If you’re looking for Instagram-type growth rates, you don’t understand the industry.
Some people suspect that Hipmunk has to spend a lot of money to acquire users. Are you on a treadmill of buying Google AdWords in order to keep traffic high?
We’re buying virtually no AdWords. AdWords make up less than 1% of our traffic. The idea that paid web marketing is the main source of our traffic is just ridiculous.
The vast majority of our traffic continues to come from organic, direct navigation.
Hipmunk launched with flight search. Then it moved into accommodation. Yet many people still associate Hipmunk with flights.
If you start in one domain, your earliest users associate you with that one domain. But is that where you make most of your revenue? Not necessarily.
In fact, flight commissions, which was our only business three years ago, now account for less than half of our overall revenue. So we’re seeing hotel commissions and we’ve added ads to the site — that combined has overtaken flight commissions as our largest line of business.
Yet flights still account for most of the searches done on our site, just as is true with most major online travel companies. But if you look at the split between flights and hotels in terms of how many are searching for one or the other, our split is comparable to much more mature metasearch companies like Kayak.
So yes, flights are still where we get most of our interactions. But in terms of monetization and revenue and overall adoption, our profile looks much more like established metasearch companies.
We don’t charge on a per-click basis. We charge on a per-booking basis — either a flat rate or a percentage. That aligns our interests’ with our partners in delivering qualified leads.
Third-party traffic monitoring services like Compete suggest Hipmunk isn’t growing. Are Hipmunk’s internal numbers on traffic and revenue more positive than what’s publicly available?
Oh yeah. Our web browser traffic is going up. It doesn’t look like those charts.
The weird thing about those sites is that they’re just not that accurate in measuring anything, period. For example, so far as I can tell, most of them don’t even try to measure mobile engagement at all, which I expect will count for the majority of our traffic by later this year.
How does the consumer shift to mobile change your business?
There’s a traditional split between flights and hotels, which follows something like an 80/20 rule or whatever for search volume, flights to hotels. But you can just throw that out the window when it comes to mobile.
More than a third of our searches on mobile are coming for hotels. And that proportion has been increasing over time, moving more in the hotel direction. And obviously hotel commissions are much higher than flight commissions.
Our mobile growth rates have been faster than any of the OTAs or metas. But we don’t want to lose sight of the overall strategy. At the end of the day, mobile is just a platform, just a way of reaching people, just like Google Glass or smart watches.
One rap against Hipmunk as a company is that it’s essentially offering commodity content. Unlike, say, HotelTonight, Uber, or Rome2Rio, you’re not offering anything that the typical consumer hasn’t seen on an OTA or metasearch site for years — except for your user experience (UX). Thoughts?
From our surveys of users coming to our products for the first time and we do focus groups and sentiment in social media and press media mentions, people say it’s a better experience.
People love the agony rating [in which flight results are filtered according to factors like price, duration, and number of stops.] People love the ecstasy rating [in which hotel listings are sorted by a combination of quality, rates, location to relevant destination hotspots, and reviews from sources like TripAdvisor]
Users also love that we show all these alternatives like Amtrak and Airbnb.
We’ve never claimed that we have the best experience for every last person, which I think is the claim made by a lot of the mainstream publicly held companies.
But for most users, Hipmunk provides an easier and faster experience. That’s something all of our research has borne out. Our challenge is to get the product in front of more people. If they see it, they like it.
What differentiates Hipmunk?
We think of things not just about mechanisms but in terms of solving people’s problem of how to book travel more easily. We don’t get too caught up in the theme-of-the-day.
If you look at the progression of travel companies, if you look at the ones that have grown to become billion dollar businesses, each of them has solved a problem better than the one before it. Expedia made it possible to book online. Kayak made it possible to compare. Booking.com made it possible book hotels you didn’t even know existed.
We view ourselves as the next step on that continuum. The user experience is vital.
Too many established brands have lost sight of their initial promise to help users solve their problems. They’ve gotten distracted by quarterly demands or the buzzy concept of the day. So our 2014 goal is not to lose sight of solving people’s core problem of booking travel faster, easier and more confidently.
Another question on mobile: Hipmunk has risen swiftly in the Apple App Store rankings in the past couple of months, according to app marketing tools provider MobileDevHQ. Has that change boosted your downloads?
We can’t be completely sure but everything we’ve seen suggests that it doesn’t matter. People aren’t searching, “travel and places” and finding Hipmunk. They’re searching for “Hipmunk” specifically to download our apps. Users hear about us by name from either friends or the press.
We also use a bit of advertising. If you’ve been browsing on your phone, you may have seen an invitation to come and try one of our mobile apps.
In summer 2012, Hipmunk made a big push for a premium $10-a-month service for businesses needing help with booking travel, building on its Google Calendar integration. Yet we no longer see Hipmunk for Business as a link on the site homepage or mentioned anywhere in the iPad app. Is that product de-emphasized now, and why?
Yes and no. We are not actively working on it at the moment. It’s something that will be a small piece of the company going forward. But what we’ve learned in doing it has informed a lot of our subsequent product development.
It has value through its features and audience, and we think you’ll see it come up elsewhere.
Will you be adding rental cars to your metasearch products soon?
It’s on the product road map at some point, but not in the next quarter for sure.
Will Hipmunk stay focused on the US and Canada in 2014, or is an international expansion in the works?
No. That said, we’ve built up lot of hotel inventory outside of US for Tonight deals and for regular advance hotel listings. Our international inventory stacks up quite well against our competitors. But we’re continuing to focus our outreach and business development on North America because there’s still enormous opportunity here.
Other metasearch companies like Momondo or Skyscanner like to quote how much transaction volume they pass along. How is Hipmunk doing?
We haven’t shared any of our top-line revenue numbers. The partners that we work with us tell us that the traffic we send them is exceedingly qualified. The customers we refer are really ready to buy.
We give our customers confidence they’ve gotten the best deal possible, not just by searching from multiple sources but also searching the kinds of inventory that others don’t, like Amtrak and private jet options next to flights or Airbnb and HomeAway listings next to hotels.
Do expectations from investors change now that Hipmunk is no longer a baby?
We’ve always been the ones at Hipmunk, not our investors, who have been driving expectations. Our expectations continue to rise. Are our expectations for 2014 even higher? Yeah. And our investors are aligned with us.
Some critics say that consumer-facing start-ups should focus all of their energies on mobile, given that that’s where the growth is and you don’t need to worry about web marketing to acquire users. The danger is that, otherwise, one is trying to optimize across too many platforms and product iteration slows down relative to the competition. What say you?
A lot of people have been talking about the virtues of being mobile-only. We fundamentally disagree with that. Mobile is an extremely important piece of travel picture but focusing exclusively on it would be a mistake for a company like us.
The use case in which mobile is comparable or better than the Web is last-minute. When you’re using your mobile and searching for hotels in particular, you will get better deals than when you were searching on the Web. Hotels are willing to discount for that perishable inventory in the final hours.
But if you’re looking for the family vacation or any of those high-ticket value items in hotels, those are things people are not booking on their phones. They’re comparing a variety of options. A tablet can kind of support it, but it’s still not the same.
If you want to capture all the value from those high-ticket value transactions as well as lower-ticket value last-minute bookings, you have to be credible on all platforms.
Many more users search for flights and then hotels, right? Most people are primarily searching for flights on the Web. It will be a long time before most flight bookings come from mobile. So what that means is that if you want to have an ongoing relationship with the customer, you need to support both Web and mobile.
Plus, we have to compete against the established brands, and it’s no secret that companies like Expedia are on all platforms. If they didn’t exist, maybe we could be mobile-only, but in order to compete we have to provide the best experience on all platforms. Mobile first is fine, but mobile-only isn’t for us.
How will the unbundling of airfares affect metasearch? Airlines want to move away from having their product be sold as generic commodities. They want to sell tickets as “custom” offers of flights paired with various ancillaries and premium services?
It’s a great opportunity. Whether or not the airlines want to share this information more broadly, and regardless of how they want to share this content more broadly, ultimately the best organizations to sell those other things are the airlines themselves.
So metasearch, by virtue of the fact that we link customers directly to supplier sites, are better positioned than OTAs are to help those airlines sell those ancillary products to relevant users at the relevant times.
We are eager to sell ancillaries. We were the first travel site to prominently show which flights have wi-fi.
But when you go down the laundry list of things that airlines are now offering, there’s always going to be a lag between what the airlines are selling themselves and what they’re making available for sale via third-parties.
It will be just simpler and easier for airlines over time to just sell it themselves — and probably more effective. Does that mean we wouldn’t show the stuff ourselves? No. We absolutely discuss with our airline partners which ancillaries we’ll show and under what circumstances and when.
But in terms of their bottom line which is in actually selling the stuff, they’re going to be more effective than anyone else at doing that.
But doesn’t metasearch currently benefit OTAs more than suppliers now?
I disagree. It depends on whether you’re looking at flights or hotels.
For flights, suppliers are winning all day long. It’s mainly about supply versus demand. On many metasearch sites, you can *only* book direct through an airline. There’s no OTA listed.
For hotels, it’s a totally different story, right? OTAs have been extremely aggressive in what they’re willing to pay to be listed in the top three positions in results on metasearch sites. Even worse, most non-chain hotels worldwide are essentially not competing in web marketing against the OTAs.
It’s hard to predict what will happen on hotels. It’s totally possible that some hotel suppliers, especially chains, will try to beat the OTAs at their own game. They may also offer benefits to customers who book direct that aren’t available otherwise.
But it’s also possible that hotels may decide it’s in their longer term interest — it’s OK — for the OTAs to be the one selling their product. As long as they can work out reasonable economics vis-a-vis the OTAs, they can leave it to them to handle the sales.
I don’t think there’s necessarily a right strategy for everyone. There will always be exceptions. For some chains brand is paramount, and they don’t want bookings via OTAs.
What about latency issues for metasearch. Does it worry you? Sometimes Hipmunk quotes a fare or hotel rate and when the user clicks through to buy that price or particular product isn’t there.
It is a problem that every company faces and every metasearch site has it a little bit worse. When you go to an airline website, when you click on a fare, most of the time they take that option out of the reservation and put a hold on it while waiting for you to decide so that other customers don’t see it for a few minutes.
But when you do metasearch, that doesn’t happen. So a ticket or hotel room may be sold in the time between when you search and when you click through to book.
And occasionally sometimes the technology is just glitchy. Sometimes an a fare shows up that shouldn’t be there. That’s happened a few times with a few airlines in the past few months.
But our error rate is comparable to industry norms. Which is to say there’s some level above which you cannot be accurate given the complexities of the industry.
You’re on a hiring streak. You’re especially looking for engineers. Kayak’s Paul English says hiring is the most crucial thing for an executive team to worry about and Paul Graham says that the difference in effectiveness between a good engineer and a great one is logarithmic. What does Hipmunk look for in a good engineer?
Hipmunk is hiring, yes. My co-founder Steve Huffman and I are engineers by background, and we take a particular and rigorous approach to hiring of engineers in particular. We’re looking for people who have deep curiosity and deep understanding of how things fit together.
There are different kinds of engineers. Some focus on one particular language, for instance.
That’s not our template. We’re looking for people who are serious enough to learn new languages in their spare time, who are working on personal projects to test their understanding, who have an extreme attention to detail — they should be sweating the small stuff to make sure code works reliably and quickly.
There’s not just one trait we look for. But if I had to only one, it would be attention to detail.
We’re looking for people who are not needing to be taught how to think about things but are coming in with a worldview from experience, which can be demonstrated by interesting projects that they’ve done either personally or at a company.
Hipmunk has 20,000 followers on Twitter, 166,000 “likes” on Facebook, and more than a million followers on Google Plus. But it takes money to run the company’s social media channels. Is the effort worth it? You offer flight deal alerts newsletters and run a blog aimed at consumers.
Absolutely. It’s part of our overall strategy. Most of our customers come in for free, so that holds down our blended cost of customer acquisition, which I think is lower than any other OTA, metasearch site, or booking app.
Has building a startup been harder than anticipated?
When we started, we didn’t know anything about the industry at all. Even as something as simple as that hotels are more profitable than flights, that was a revelation to us.
On the other hand, the components to building a startup that are common across all verticals, those are things we knew would be hard. We’ve both been involved in startups before.
We knew we’d have to keep hiring standards high, number one. We knew it would be hard to go through that first painful moment where you have to let someone go. We knew what to expect when it came to managing people, coordinating big projects — all those kinds of things.
But when it comes to understanding the dynamics of travel industry, that’s was quite hard, and also really time consuming.
Suppliers have a love-hate relationship with intermediaries, and OTAs have a love-hate relationship with them, too. As a metasearch company, we’re in a position where we have to be very aware of all the dynamics around us.
Each company’s strategy, and the power dynamics, and keeping all that straight and keeping everyone satisfied and at the same time making sure that we’re not compromising our users’ experience in ways to satisfy suppliers’s most assertive demands has been a very tricky challenge.
And I don’t see that getting any easier. It may get even harder as more divergent strategies on distribution are pursued.