Steve Kaufer, Langley Steinert, Nick Shanny, and
Thomas Palka started TripAdvisor, an online travel
site, in 2000. Frustrated by the lack of unbiased, useful
information for travelers, they created a site that,
in addition to searching relevant content already on
the Web, let users contribute personal reviews of destinations,
hotels, and attractions. The online travel
forum was a pioneer in the now common practice of
having users pick the winners, instead of leaving the
choices up to human editors.
TripAdvisor became the largest online travel community
in the world, and was acquired in 2004 by Barry Diller's InterActiveCorp
(IAC). As of July 2006, TripAdvisor had amassed more than five million user
reviews and opinions, covering 220,000-plus hotels and attractions.
Livingston: How did TripAdvisor begin?
Kaufer: The idea came when my wife, Caroline, and I were trying to find a
vacation for ourselves. We started with a travel agent, who recommended an
island and some resort. This was '98 or '99, and I thought I'd use the Internet to
find out more. I found a whole lot of websites that would help me book a reservation
at this hotel, but nothing that would tell me whether the hotel was any
good or not for what I was looking for.
Eventually I found some chat rooms that told me that the island was not
particularly safe, and we really didn't want to go there. That was kind of an eyeopener.
We said, "Good power of the Internet there. Let's switch travel agents."
We went to a different one, who recommended a different island, different
place. That time, when I did the research on the Web, after a lot of effort I
found out that the hotel was really not up to my wife's standards. The picture of
the hotel in the brochure was fabulous, of course, but the commentary from
somebody's home page that I had found wasn't too good.
361
Stephen Kaufer
Cofounder,TripAdvisor
26
CHAPTER
By that point, I had spent a couple of days in sort of mindless searches, trying
to find the real scoop on the hotel, not the official blurb. My wife suggested,
"You know something about technology. You could build a better search engine
to find what you're looking for in travel--not the published opinion, but the
unpublished, unbiased opinion about a place, a location, something to do."
I was employed at the time, so we put the idea on ice for about a year. In
late '99, the idea resurfaced. I wanted to get out of what I was doing, and
started to assemble friends that I had worked with before who might be interested
in starting an Internet company to build the best travel search engine out
there--where we would define "best" as not searching for prices, but really
finding the unbiased information.
I was introduced by a friend to another cofounder, Langley Steinert, on the
business, marketing, business development, financing side of things. So the two
of us kind of took up the project as, "Hey, this is something the world clearly
needs." I felt I could build it with the team of folks I had in mind from past
lives. Langley had the business development experience and connections to sell
and market it. Because I had started a few companies before, I knew it was
important to have the right combination of skills and interests amongst the
founders. We assembled four initial founders of the company and got our first
round of funding in February of 2000.
Livingston: Where was your office when you started?
Kaufer: My late wife actually owned a software company that was just down the
road in Needham [Massachusetts]. It was a small and declining company,
which, for the first 10 months or so of our existence, gave TripAdvisor free rent,
T1, computers, and other stuff that it had and wasn't using. So it wasn't technically
a garage. It was closer to a second-floor attic above a pizza place. It was all
one big, open floor, and the room could comfortably seat eight. By the time we
busted out of there, we had 15 people. Then we just moved down the street.
Livingston: So your idea was to somehow collect the consumer feedback on
different hotels, airlines--anything related to travel?
Kaufer: We were going to focus on destinations, hotels, and attractions. We've
always pretty much stayed away from collecting opinions on air, for instance.
But we were going to search the Web, just like Google--or AltaVista, which was
king of the hill in those days--but with a focus on travel. We'd be able to come
up with better results, where better was, again, not just all the booking sites that
would help you book a room in a hotel, but really opinionated information.
We'd find the articles from the New York Times, Boston Globe, LA Times,
local newspapers, etc. The back issue of Ski Magazine might have a great article
all about Aspen, but you'd never find it, because it's tucked away in the archive
section and probably wouldn't show up on Google. It was written last year--
"Great Things To Do for Families in Aspen." A fantastic article, and what was
good last year is probably still just as good today, but you'd never find it but for
our very focused travel search engine.
Livingston: How was the technology designed to do this? Was it a crawler?
362 Founders at Work
Kaufer: We tried different approaches. We tried to randomly crawl the Web
and we thought, "How are we going to randomly select out of the billions of
pages?" So we tried crawling from known travel hubs. We'd start from the
Yahoo Travel directory and see where those sites led us. We tried to pick out
good, interesting information and automatically categorize it. That didn't work
so well. What we call the signal-to-noise ratio wasn't good enough--meaning
that, when they got our results back, people wouldn't say, "Oh yeah, that's what
I was looking for."
We ended up looking at all of the published sources of information--newspapers,
magazines--and manually went through all the websites from all these
places to find the ones that had free access to the back issues of their travel articles.
Then we hired people to read every single travel article we could find on
the Net, and classify that article into our database, and write a one-line summary.
It's a fairly significant effort, and people that we talked to said, "You're
nuts. You'll never finish." But if you actually do the math, you realize that you
can work through the backlog (it took us a couple of years, but it was only a
couple of years) and then can stay current with what's being published without
too much of an effort.
We take half an hour to read an article, on average, and we'll tag that article
as being relevant to everything the article talks about. If the article is about
Maui and things to do in Hawaii and these two resorts, whenever you're searching
for Maui or things to do in Hawaii or those two resorts, that article will
come up. If that article happened to mention, "The beaches in Maui are much
better than the beaches in Fort Lauderdale," and you were to search on the
beaches in Fort Lauderdale, that article is not going to come up, because our
search isn't keyword-based. It doesn't matter if the article happens to mention
something; you only want to read the article if it's actually giving you an opinion
on the topic you're researching.
What we ended up with was a much smaller database as measured by the
number of documents that we'd indexed, but extremely, extremely relevant.
You go to a page about Maui, and every article on that page really is about
Maui, sorted to a pretty good degree based upon which article most people
would rather read first. Would you rather read an article that has a paragraph
about Maui in talking about fun beaches around the world, or an article all
about beaches in Maui? Probably the latter, so that's why the article is sorted
first. Your experience on TripAdvisor--again, this was initially, when we
launched the site--was very fulfilling, because the information we found was
always spot-on. We didn't always have something, but what we had was always a
match.
Jumping forward in time as the site grew, all of a sudden now those hundreds
of thousands of articles are dwarfed by the user reviews that our visitors
have generated. It's fresher information and tends to be more detailed. To
many people, it's more reliable.
There's a whole other theoretical discussion of, "Would you rather read a
review about a hotel by Aunt Mary you've never heard of from Bloomingdale,
Indiana, or from Frommer's, the trusted guidebook brand?" And the follow-up
Stephen Kaufer 363
question is, "Would you rather read 20 reviews from people you don't know, or
1 review from Frommer's?" Near as I can tell, most people, when given the
choice of only one piece of information, will take the Frommer's--even though
they might be suspicious it's a little old or a little vague. But when you have 10
or 20 reviews, and you have a half a dozen written in the past couple of weeks,
you know you're getting an unvarnished and up-to-date version of what you're
looking for. And colorful.
Livingston: Were the people originally gathering all this content TripAdvisor
employees, or were they contractors?
Kaufer: A combination.
Livingston: You said it took a couple years to populate the site. Did you launch
the site before it was fully populated?
Kaufer: Oh yeah. We started in February 2000, and in October 2000 we
launched the site, but it only covered the United States. Over the course of the
next 2 years, we rolled out the rest of the world geographically. Of course, we
were always adding more and more content as we found it. When we launched,
if you picked the 20th hotel in our popularity index in Boston, there might have
been one or two articles about that hotel, which is a heck of a lot better than
none, but nothing compared to what we have now.
Livingston: How did people find TripAdvisor when you first launched?
Kaufer: When we started TripAdvisor, the notion was TripAdvisor.com was
actually just going to be our demo site, because we never planned to appeal
directly to end users. We were going to be selling this rich database to travel
portals, online travel sites. They would be querying our database to find the
best information and surfacing it to their users, and there would be a little
"Powered by TripAdvisor."
Because we would have the richest database of travel information, our hope
was that it would become a requirement that, if you were in the travel industry
or offering a travel section of your site, you have access to our content. And we
would license it out and/or get a share of the revenue generated on the page
views from that. Lycos Travel, Yahoo Travel, AOL Travel, Expedia, Travelocity--
all the players would have to have it. No one would try to build it themselves,
because we'd always be able to stay ahead, since we were entirely focused on it.
That would be our business model, and that's the model that got us some funding
to begin with.
After a year and a half, we had closed one licensing deal, with Lycos, where
they were featuring our content on their travel portal, and we were getting a
revenue share on what they made selling advertising on the pages that we produced
for them. Everyone else basically wanted to be paid to feature our content,
and we wanted to get paid to have our content featured. So there was a
pretty big disconnect. Then it turned out that with the Lycos deal (even though
Lycos was a major web property at the time), the joke was the quarterly
revenue check wouldn't buy the weekly free lunch that we offered to our
employees. We had a rather fundamental problem in half of the business. It was
364 Founders at Work
a typical dot-com business problem: built the product, people liked it, and the
feedback was universally positive, and we got the expansion questions that we
were happy to get, like "When are you going to cover Paris?"--but we just were
not making a dime.
By the middle of 2001, we were getting frustrated. Then September 11
came along, and anything we might have had in the pipeline--not that I
remember it being a particularly interesting pipeline--was stalled, dead. It was
a hugely traumatic time for everyone, especially for the travel industry.
We were also trying to raise a third round of funding, and we were basically
looking at going out of business in 6 to 9 months. It was a little hard to go back
to the existing investors and say, "Hey, pony up more money. We've got a great
product. We have no revenue, and we've been trying to sell the stuff for a while.
We have no takers. The one company that did license it is generating a couple
hundred dollars a quarter for us. But really, toss in a couple more million,
because it's a great idea." It was a tough pitch to make. We made it, and we
actually did raise a small third round--more, I think, from the perspective of
"Look, this is a good product. We'll figure this out."
We were 11 people before September 11, and we slimmed down to 8, so
our burn rate was really pretty small. Everyone took salary cuts; we were paying
$18 a square foot for office space; we had really no expenses to speak of. We
were stretching the dollars, and even though it was Internet dot-com days, we
had never done anything remotely lavish.
So we're approaching late 2001, and we noticed our demo site,
TripAdvisor.com, had started to get some traffic. Just people finding it. We tried
to be active in PR from day one, so we'd gotten some mentions in various press.
I'm not entirely sure how people were finding it--search engines, whatever.
And we certainly weren't doing anything monetizing the traffic. We thought,
"OK, with 5,000 visitors a day, let's go run some banner ads, see if we can make
some money that way."
We tried running a banner ad. We didn't try to sell it; we just copied
Expedia's banner ad and put it up on our site. We wanted to see how many
people would click on it. We might have had 3,000 visitors that day and we
might have generated 100 clicks, so maybe it would have been a couple of
dollars to us. So that was just clearly not going to work. But one of our prospects
a couple of months earlier had asked us whether or not we could run ads based
upon the search query. If someone was searching on "Boston," could we run an
ad for Boston? We explained, "We don't run ads. That's not our model. We're
trying to license you the content." But it struck us months later that we do have
people that are qualifying themselves to be interested in Boston. In fact, we
have people qualifying themselves to be interested in the Eliot Hotel in Boston,
because they're reading a review about it. What if we created a link from
TripAdvisor deep to an online travel site like Expedia and had teaser text that
said, "Book a room at the Eliot Hotel in Boston," and, if the user clicked on that
link, we took them all the way down to the booking page on Expedia? Our
crawler technology knew how to do that, so it was leveraging something we
were pretty good at.
Stephen Kaufer 365
We approached Expedia and said, "Hey, we'd like to advertise your 50,000
hotels on our comprehensive travel site, and we want to charge you only for the
leads that we'll send you." I explained how our leads were highly qualified.
"These are wonderful travelers, they're reading reviews, and we're going to do
lots of bookings for you, so we'd like you to advertise with us." He said, "I've
never heard of you guys. Prove it."
I can certainly understand why he'd never heard of us, because no one had.
And he didn't really care much whether I was sending him qualified leads or
not. He only cared about whether the leads I sent were going to make reservations.
The way it works is, they give you a tracking code to put on the link. We
said, "It's no work for you. Give me the tracking code; I'll start to advertise
50,000 properties for you." Which I did, and we ran it for a month for free for
them. Then, before the month was out, I gave them a call and said, "How's it
going?" And he said, "Well, I just looked at the stats the other day. You guys are
doing pretty well. Can I pay you $10,000 for December to buy 20,000 (or whatever
the exact numbers were) leads?" Here was a client who was on the second
or third call offering to pay me money to keep the links up.
This worked. We actually got a lot of people clicking. Probably 10 percent of
the time that people saw that page, they were clicking on one of those links.
Click-through rates at the time were a quarter of a percent or half a percent.
Here we were sitting at 10 percent because the links were so relevant to the
topic at hand. Our first client was thrilled with it.
Livingston: Expedia was the first client?
Kaufer: Yes. He said, "If they stop actually doing bookings and they just click
over, then I'm not renewing the order. So it's up to you guys to keep your traffic
qualified." This wasn't really an issue for us, but it just drove home the point: if
he got bookings, he would happily pay us for advertising.
We said, "If we do more traffic in January, are you willing to spend more?"
He goes, "Yeah, we can ramp it up to $20,000 if you have more leads." Once he
got a little comfortable with us that he wasn't going to lose his job over committing
to a faulty buy, it quickly became, "Hey, send us as much as you can.
There's no cap on this number." And the guy that I was starting with--talking at
ten grand a month, we were in the hundreds of thousands not too much later--
it would be another year before I actually met him, because it was just a very
practical, "You're sending me leads that I'm tracking through to people that
actually buy products on my site. I know how much I make when they buy my
airline ticket or my hotel room, and I'm paying you a percentage of that. So the
more leads you send, the more money I make."
TripAdvisor never knew whether we were earning 25 percent of their profits
or 90 percent of their profits--the answer is somewhere in between--but it
became a pretty easy sales cycle, if you will. And then we started growing traffic
to TripAdvisor.com, and we started expanding our client set beyond Expedia, to
hotels.com, Travelocity, and eventually Orbitz and others.
366 Founders at Work
Livingston: Did you use the same strategy with other companies: "Try us out
for a month, and, if you feel like we've driven some true leads, you'll continue
with us"?
Kaufer: Yes. Once other companies saw Expedia advertising, they sometimes
didn't need a free test, but we might say, "Look, our leads are normally a dollar
a click, because they convert so well. But we'll let you get started at a quarter.
And we'll send you 5,000 leads, and you can test with no risk. But you know,
we're looking for an insertion order to show that you're committed to the test."
Invariably it would take three months to get a test going. You have to find
the right person; you have to introduce yourself; you have to decide whether it's
an ad agency or direct with the client and lots of other annoying aspects. But for
the most part, once the client was getting the leads, the leads would convert
well enough such that they would be up and stay up for years and years. So
there wasn't a whole lot of maintenance involved. Our technology would automatically
find the right links to advertise.
Livingston: It sounds like finally figuring out how you were going to make
money was a major turning point for you.
Kaufer: Right. We went from no revenue to break-even in the course of about
4 months. That part was a testament to finding a model that worked. To break
even, I had to do $75,000 in revenue for the month, something like that. We
had never let our burn rate grow. We didn't do any advertising at that time. But
even since, we're rarely going to do promotions that we can't tie back to actual
revenue-generating activities on the site.
Livingston: Why were you so careful about spending money? Had you had a
bad experience before?
Kaufer: By 2000, we'd certainly seen the dot-coms that would move into the
$50-a-square-foot offices, hire loads of people to get it all done quick--to get
big fast, etc. Several of them had already flamed out. That was really never in
my blood, if you will. The other company I had started right out of college was
self-funded, and then a tiny bit of angel investing, and then half a million dollars,
and then a couple of million dollars, where the last round was purely
growth capital. We had always run that profitably and had grown slowly because
of it.
I guess I had toyed with the idea of doing the same here--taking my savings,
building the product, not looking for any venture money--but that would
have taken a long time. I had a family to support, that sort of stuff. So we certainly
decided early on to go raise some money, and within the first year we had
raised $3 million or so--but a comparatively small amount of money.
There wasn't anything obvious that we should spend money on, other than
hiring a lot of people, and I'm just a fundamental believer in small teams do
better than big teams. We were building a product, and if there were 5 people
and they were all within shouting distance of each other, they were going to
build a better product than if we had had 15 people. Really, even if I had the
dollars, I didn't want to spend the money there. And then, did we need more
Stephen Kaufer 367
sales reps without a product to sell? No. Could we have bought more PR? I'm
sure we could have. We actually had a full-time PR person pretty early on. But
we were never going to raise $20 million for a marketing blitz.
Livingston: When you first started trying to get customers, they said, "Pay you?
You need to pay us." What were some of the other things that went wrong when
you were trying to figure out your business model?
Kaufer: Building the product actually went along reasonably well. The sales
and marketing and business development was the biggest challenge, because
we just didn't have any takers. We finally had a major company come forward
who wanted to license our database, and they were offering to pay us, I think,
$50,000 a month. Maybe it was $30,000. The deal on the table would have
covered half of our burn rate, right there, all at once. At the time, we looked at
that deal and said, "You know, it may be this or nothing"--as in, if we don't take
this, we might just go out of business. When you look at a deal in that light, a
very unattractive deal, maybe that's what you have to do to survive.
As we proceeded down the negotiation path with this company, though, it
became clearer and clearer that they wanted to be able to cut out of the deal
after the term (I think it was 2 years) and essentially walk away with all of our
intellectual property. Their point was, "Hey, our dollars are funding a lot of the
creation of this thing. We're going to be building a much bigger product around
it. If we can't renegotiate terms, you can't cut our product off at the knees and
walk away with your database. We need to be protected, and we need a copy of
your database and the tools to maintain it. You won't have to maintain it for us
anymore, but we'll be able to keep going."
The notion that I take all of my time and energy, build up a business, and
then hand another company who is going to be a competitor the crown jewels
of the business--fundamentally the business, except for the people--after I
thought about it, that ended up making it a reasonably easy decision. "No. I'd
rather go out of business than take everything I'd worked for, for so long, and
hand it essentially for free to somebody else."
In hindsight, we clearly made the right decision. But at the time it wasn't
obvious, so we kept negotiating with them. My tip for someone in a similar
situation--a company looks like it might be going out of business--this might
be the only way to get enough capital to survive, on terms that really aren't very
acceptable: keep pushing at it. Don't say yes too quickly, don't say no too
quickly, to see whether any other options come along, or to see whether the
deal gets so bad that it actually becomes an easier decision to just say no and
you can go about your plans. We had a few sleepless nights back then.
Livingston: Can you think of a moment when you wanted to quit?
Kaufer: No, I never wanted to quit. I mean, it wasn't working, but I was going
to find a different way to try--something different, something new.
A good chunk of our engineering team was directed to running tests for
prospects who might conceivably be clients. We weren't doing a whole lot of
new product innovation, because it wasn't the product that was stopping us
368 Founders at Work
from being successful; it was the sales and marketing strategy. I said, "Look, we
will take everyone in the company and turn them into somebody that's going to
help close the sale to keep this company afloat. So you're doing one of two
things: you're either helping on some prospect we were chasing, or you're helping
in the financing prospect we were chasing."
I was always certainly cognizant of the fact that, if we didn't do something
different, we were going to run out of money in the first half of '02. But it never
crossed my mind to just give it up or shut it down.
Livingston: A lot of startups that are based strongly on technology don't have
the luxury of having a business guy as one of the founders. Do you think that
having Langley on your founding team helped you?
Kaufer: Absolutely. We never would have succeeded without Langley on the
team. If I were funding a startup, I wouldn't want to put money in unless I saw
somebody identified as having an interest in the business development side of
it. I'm an engineer by training myself, but at this point I have so much experience
dealing with customers and what they want that I can bring that back to
shape the product. Look at me 20 years ago, and at best I was a smart engineer.
I didn't know much about business, knew nothing about selling, and unless you
have somebody who has an interest in talking with whoever you're selling your
product or service to, your product isn't going to turn out to be what the customer
wants.
In almost all circumstances I can think of, if not a member of the founding
team, you want to say, "With the money I hope to raise from you, this is the person--here's
his/her resume--that we're going to bring on board to take care of
the business marketing aspect of it." I've always had that in the startup companies
I've been associated with.
Livingston: What competitors were you most scared of as you were building
TripAdvisor?
Kaufer: There weren't really direct competitors. We were fighting more of a
problem of, "No one else is using your stuff. We seem to be doing OK without
it. So why is your stuff critical? Why do I have to pay you for it?" The dollars
that might have been spent on us were probably going to a Frommer's or
Fodor's, which were branded content sites.
We would say, "No, no, no. They have one person's opinion, and it was written
6 years ago by someone that may or may not have even visited the hotel.
We've got fresh stuff. We've searched for all the stuff around the Web that your
visitors want."
But, there were drawbacks to our model, too. A user on Yahoo Travel looking
at their description of Boston was reading it on Yahoo Travel. When they
came to our Boston page and they wanted to read about fun things to do, we'd
take them off to an article on the New York Times, or on Frommer's, whereupon
they would be leaving TripAdvisor or leaving the Yahoo network. Yahoo, like
most companies, didn't really want to send a lot of people away. Yet that was
how we had such a rich database of information. So a tougher sell.
Stephen Kaufer 369
Livingston: What other things did your customers misunderstand about
TripAdvisor, since it was the only one of its kind?
Kaufer: I don't know that they misunderstood too much. Today, user reviews are
in many spaces a matter of course. Amazon has done a tremendous job turning
that into a significant competitive advantage. In 2001 or 2000, Yahoo wanted to
get more than (I'm making up the numbers) $20 million dollars for a 3-year
Travelocity contract; they wanted to get $40 million dollars. They wanted to sell
a sponsorship to Carnival Cruise Lines for a million dollars. They wanted to
monetize their existing travel channel as opposed to improve the content by
licensing with TripAdvisor.
It wasn't until around 2004 that we looked at Yahoo Travel as a competitor.
It didn't change in 2000; it didn't change in 2001. I mean, the ads changed.
They got better at extracting more money, I guess. But the actual content--the
reason to go there--didn't change for 3 years straight. It was great to have them
as a competitor, in the sense of, you know, pathetic. In 2004 they said, "Whoa!"
and really made dramatic improvements, and they built a much better product.
Livingston: Did your writers edit the users' feedback submissions?
Kaufer: No.
Livingston: How did you monitor the entries? Could people say, "This place
sucked! Don't go there."
Kaufer: Yes. "This place sucked. Don't go there. Found rats under the bed."
Very colorful comments on all sorts of stuff. We'd frequently get threatened by
hotel operators who were unhappy with the reviews that were posted.
We look at all the user reviews that come in every day and make sure they
meet our posting guidelines: Is it family friendly? Are you using hate speech? Is
there racist commentary? We will not edit the reviews at all. We'll either reject
a review or allow it to be posted.
Sometimes we make mistakes and post stuff we shouldn't. But those are the
mistakes. We'll let the horrible reviews come in and, obviously, post the great
reviews as well. If a hotel complains, "Hey, this person is lying about my property.
They never stayed here. We never had any record of anyone staying here,
blah, blah, blah," we say, "We have a form on the site where the hotel management
can post a response, so that our visitors can see both sides of the story."
But we won't take down a posting if a hotel owner complains. And we make no
attempt to verify the factual accuracy of a review. From our perspective, we
have to be a little concerned about libel laws, and we fall under the sort of communications
act that says we're a conduit for consumers talking on the Web.
You can't sue AT&T for hate speech said on the phone line that they own. You
can't sue TripAdvisor for libelous statements that appear on user reviews.
Livingston: Can you think of one example where someone wrote a really
scathing review, and the hotel got mad at TripAdvisor?
Kaufer: There was a hotel owner in Italy who had their attorney draft a letter to
us, actually all in Italian, saying essentially, "If you don't take this review down,
370 Founders at Work
we're going to sue your butts for $2 million." We sent it over to our legal department
so they were aware of it, but our response was a polite, "We verified that
the review meets our guidelines." End of story.
In 5 years, we haven't been sued by anyone--because, when you actually go
look up the law, we're protected. There may be some hotel owners with a legitimate
beef, but there are some hotel owners that are really just not very bright.
They'll complain about how terrible this review is and their email is from
abc@yahoo.com, and then the next day a review will appear on their property,
written by an email address of abc@yahoo.com, saying, "I stayed at this place,
and it was absolutely magnificent. The views were spectacular." I'm thinking,
"You're just writing a review, posing as a guest, on your own property--24 hours
ago you used the same email address to tell us that you were the owner complaining
about a past review. Do you really think we're that dumb?"
Initially it really didn't matter. Our traffic is so high now that we know, for
better or for worse, we have a significant impact on where visitors are choosing
to stay. For every city, we kind of have a satisfaction index; we rate which hotels
our travelers like the most. If you're ranked first or you're ranked 20th, the
number of reservation calls or bookings you're going to get is going to change.
When we changed our algorithm, it dropped some hotels and raised others.
Our phones were ringing, because we had had a material effect on their
businesses.
When Google changes their algorithm today or when they update things, it
has a significant effect on the people who run their business based upon getting
traffic from the Google search engine. For us it's a responsibility, because we
want people to trust TripAdvisor. People absolutely post scathing reviews. But
we don't want to be spammed. We don't want hotel owners to tell all of their
employees to go write wonderful reviews of the property. So we have our techniques
and our human and algorithmic ways to detect that sort of fraud, to keep
the accuracy of TripAdvisor as high as we can.
Livingston: I like that your site has an impact on keeping the hotel owners
honest.
Kaufer: It's not just for traveling, but for a lot of things that people are collecting
user reviews on now. You may go to Cincinnati for a business or leisure trip.
If you have a bad experience with a Super 8 motel, before a site like TripAdvisor,
basically there was nothing you could do about it. You could file a complaint at
the Better Business Bureau, but how many times have you, when making a
reservation at a place you've never stayed at, called up the local Better Business
Bureau to find out how many complaints were lodged? None. So here you go to
TripAdvisor and you look up the place, and you see that seven out of the last
eight reviews all gave it a 1 out of 5 rating and talked about smelly carpets and
rude staff. You're just not going to stay there unless it's the only place in town.
That's the impact that five or six people had. Total strangers. But the hotel
owner that wants to run this crappy place, preying off of the brand that they're
under, and maybe their location as being near to something, that person has to
kind of shape up, maybe take something out of their profits and put it back into
Stephen Kaufer 371
providing a good service for the customers, because word is spreading. And
TripAdvisor is the conduit in the travel space for spreading that word.
Livingston: Looking back on your experience with TripAdvisor, what was most
surprising about it?
Kaufer: Certainly the most surprising to me was how much people voluntarily
share their experiences. I had never written a review before starting
TripAdvisor. I had never posted my comments on Amazon or anywhere else.
But we're able to collect millions of opinions each year. It is 2006, and we're up
to over five million now. So you have a lot of people out there that, for
absolutely no reward whatsoever--we don't pay them for opinions, we never
have--will take the time to write a review or answer someone else's question.
Maybe it's because we're the size we are and that people have gotten a lot of
content, so now they're interested in sort of giving back to the site that helped
them make a decision. But the fact that we're able to collect so much on an
ongoing basis isn't something I would have predicted.
Livingston: Is there any advice that you would give to someone who is thinking
about starting a startup?
Kaufer: Certainly the founding team makes the biggest difference. Usually
founding teams don't stay together for very long. That happens. If the founding
team splits in the first 6 months, that can be pretty devastating to the birth of a
company. So getting to know someone before actually joining forces, spending
some more time thinking through what the roles and responsibilities will be
between the founding team members. You hear breakup stories of, "Well, I was
going to do this." "No, I wanted to do that." "Oh, you're taking too much control
over this." Unfortunately, odds are high that'll happen anyway. But if you
can iron all that out before you actually start the company, or pick different
founders, it'll improve your chances of success.
Tip number two: you can't get too attached to your vision in a startup,
because things may change. It's not a sign of failure to change your vision. I
remember in a previous company, we wanted to be this, but we were offered a
consulting contract to do this, that, and the other thing, and, yeah, that wasn't in
the plan, but we'll take that, because that's going to add $50,000 to our startup
capital, and it'll only take x amount of time. Yeah, be wary of distractions, but if
you're lightning-focused on just one thing and aren't willing to consider others,
you probably don't have the flexibility to make it when things don't go according
to plan. That's the one truism: things won't go according to plan.
At the earlier stages of the company, when you're actually out trying to get
some customers, do whatever the hell the customer wants. If they're going to
pay you, the customer is right. Because you need that initial money. You need
that customer on the list to go get the next one. If you have to give away whatever
you're doing, give it away. Get the customer. Make them into a reference
account. Make that customer into the person that sings your praises the
loudest, and really uses your product or service.
372 Founders at Work
It's perfectly fine if you knew that customer through a past life and that's
how they got to be a customer. Maybe they'll be even more honest with you. If
it means adding a new feature to your product, or whatever, to close that initial
sale, and it's not on strategy, screw the strategy. Do it, collect the money, get the
customer, and move forward.
Then, as you're growing, sort of mid-stage, what I tried to foster here is an
attitude of risk-taking, where all I want to know really is what's my downside
scenario in terms of time and opportunity cost? If I can try something offbeat,
weird-sounding, and it takes a few weeks, and the downside is, in my case, I rip
it out of the live website, or, if you're producing software, I rip it out of the
piece of software, or I don't document the feature if I'm producing a piece of
hardware or whatever--if the amount of time spent making a mistake is small,
don't be afraid to make a lot of mistakes without a lot of time analyzing whether
you should or shouldn't do it.
On the Web, it's particularly easy to try something and get feedback. If it
doesn't work, drop it. I've come up with really interesting ideas that were utter
and complete failures on the site, and I make fun of myself in company meetings
when I talk about those. Then I look at each group and say, "Hey, I'm
hoping every one of you--in addition to all the successful ideas you'll come up
with--aren't afraid to come up with some resounding failures." You just want
the failure to cost you a couple of weeks, a month or two--it depends on the
industry--a small, fixed cost. It's the old adage: if we're not failing at something
on a regular basis, we're just not trying hard enough.
Livingston: Obviously your story ended wonderfully. You were acquired for
around $200 million?
Kaufer: Yeah. I don't think of the story ending that way, but that's how the third
chapter ended.
Livingston: Sorry, I didn't mean it like that. But most startups do want to have
a liquidity event. Would you have done anything differently before that, to get
there?
Kaufer: No. With TripAdvisor, it all really did work out well. Certainly one of
the keys to our success was being fanatical on the hiring side of things. I was
almost going to answer, "Well, I would have liked to have hired more top-notch
folks throughout the company earlier." Because I'm still in that position now--
I'm still struggling to fill positions with the types of people that we want to hire.
It's not something that we do very efficiently here. It takes us a long time to fill
a req.
When we do fill a req, we have a fantastic success rate. Many observers and
people that have done due diligence on TripAdvisor over the years have commented
on the caliber of individuals here. But if I ever start another company
again, I'd love to have as a founding or very early team member someone who
was a trusted recruiter. Because the difference in almost any position between
someone who does a good job and someone who does a great job might be
20 percent more in salary, but it's 100 or 200 percent more in throughput. If you
Stephen Kaufer 373
can have enough people in the company that work twice as efficiently as the
person sitting next to them, because they just know what to do, what not to
spend time on... I mean everyone, they're more or less all working the same
number of hours. It's rarely a work ethic issue. It's just, hey, you give this engineer
a task, and it's just done right in half the time as the next person. That it's
done right, that's the first important part; it's done quick; and there's just less
communication if the teams are smaller, because everyone's getting twice as
much done. Now how the heck do you fill a company with people like that in
every single department? Well, you tap out of your friends pretty quick; but
absolutely, go hire your friends.
As I advise other startups from time to time, if you find someone you like,
pay what it takes to get them to come to your company in options or in salary,
depending on the company's stage. But getting the right people--especially in
that first dozen--is so much more important than getting the req filled.
Unfortunately that slows down the hiring process a lot, which slows your
growth a lot, which is how I circle back to say, "In the next company, I'd hope to
have a recruiter on board within the first half a dozen people to help get the
right next 12 people."
Most recruiters don't work that way, don't think that way. Recruiters want to
know, "What requirements do you need in the job?" My answer is, I want passion.
I want people that really care about doing a great job. It's just a different
mindset. That's software, that's customer acquisition, that's branding, that's PR.
It's really not in any one department. It's an attitude. And it makes a company a
hell of a lot more interesting to work at.
So in turn, it actually makes recruiting a little easier, because you come in,
you meet the people, "Man, you've got a bunch of sharp people here." "Yeah,
that's right. And Expedia did a customer survey for us, and it came back that
98 percent of the people said they really enjoyed working with the other
people here."
Livingston: Does Barry Diller let you do your own thing, now that you've been
acquired? How was the transition to being part of a larger organization?
Kaufer: IAC was a fantastic company to be acquired by, because they told us
their history was, "We acquire companies and we let them run stand-alone."
So we came on board. I have to report finances up through IAC instead of my
own five to seven person board. Other than that, we were really left on our own.
Even when we had ups and downs in our numbers, we were left on our
own. They were true to their word. And my hat's off to them, because you so
often read about small, entrepreneurial companies like ours being acquired and
it being a disaster, because founders leave, teams quit, an infrastructure process
is put in place that doesn't fit. IAC said, "Hey, you're running a business. Tell us
what you're going to do, then do that. And let us know when you need help."
That's what we've done.
Livingston: I noticed you don't have a receptionist.
374 Founders at Work
Kaufer: I couldn't figure out what a receptionist would do. And executive assistants,
we don't have those either. So on the subject of hiring, I don't look at a
head count budget when I think of hiring people. I wait until I see the need for
someone--when I can carve out a job description that's 80 percent full on the
day someone starts--and that's when I'll open up a new req.
For receptionists and executive assistants, it's something of a running joke
because, well, what are they going to do? Make travel plans for us? We already
know how to do that. Answer the phones? Well, I can answer my own phone; it's
not too much of a bother. Schedule meetings? Well, we try not to have too
many meetings to schedule. So if we were to hire one, I'm sure they would be
busy all the time, but perhaps not doing anything that really needed to be done
at TripAdvisor. With engineering, there's usually more development that I want
done, but I can look and say, "Am I willing to fund that project?" With marketing,
with customer acquisition, with accounting, I first look and say, "Hey,
what's taking people's time? What can we automate?" If I can't automate it, do
I really need it to be done? If I do need it to be done, all right, then we'll open
a req.
We're 70 people now, which is pretty small, given the revenues and profits
that we're producing. Nobody in corporate would blink if I said, "Hey, I want to
have 20 more people on board." Our margins would still be terrific and I could
afford it, but I'm not sure it would speed things up or slow things down.
Livingston: It sounds like you're really maintaining a good atmosphere, and
one where innovation can happen rather than just saying, "We've been bought.
I'm leaving now. It's 5 o'clock."
Kaufer: Chapter one is starting up the company, getting anyone interested in
what we do. Chapter two was hitting that profitability mark so we could break
even. Chapter three is growing, and if it so happens, as it did with us, a great liquidity
event. Chapter four is, "OK, are you done with the business? Do you
already have a commanding number one market share and no competitors?"
Well, that's pretty rare. Certainly it isn't true for us. We may be ahead of our
competitors, or tied with our competitors, but how close am I to being the
number one travel site in the world? Well, I'm pretty far away from that. I have
some sister companies that have that spot. But I look at it and say, "Why
wouldn't everyone want to start on TripAdvisor when planning a trip?" So if I
have 20 million uniques a month now, why don't I have 50 million? My definition
of chapter four is being the most popular travel site in the world.