Utrip

» see utrip.com as it lived, on the wayback machine

~ the eulogy ~

Gilad Berenstein was 23 and had just finished graduate school at the University of Washington when he tried to plan a trip through Europe and found the work of it ridiculous. The information existed, in guidebooks and review sites and forum threads, but assembling it into a schedule that accounted for which neighborhood you were already standing in was a job nobody had automated. He started Utrip in 2011 with his father, an engineer, and a third co-founder.

The product asked you a small number of questions about what you cared about, and weighted them: more art, less nightlife, this much walking. Then it produced an actual day-by-day plan, with the ordering solved for travel time between stops. Calling it artificial intelligence in 2014 was a stretch by later standards and completely normal by the standards of the time, and the thing it was really doing was a recommendation engine with a map and a clock attached.

It came out of beta in 2014 on a $750,000 round led by the chief executive of Virtuoso, added another $850,000 with Costco's finance chief among the new investors, and by the end of that year had hundreds of thousands of users across 32 markets. Berenstein won Young Entrepreneur of the Year at the 2015 GeekWire Awards on a vote of the Pacific Northwest tech community. For a while Utrip looked like the one consumer travel startup that was going to get through.

Then the arithmetic of consumer travel caught up with it. Acquiring a traveler costs real money, a traveler plans maybe two trips a year, and the people you are bidding against for that attention are Expedia and TripAdvisor and eventually Google. Berenstein's own summary was that driving traffic is hard and expensive.

~ the dossier ~

~ the pivot ~

So Utrip stopped selling to travelers and started licensing the engine to the businesses that already had them. The customers were hotels and cruise lines, plus the convention and visitors bureaus of mid-sized American cities. The pivot worked on its own terms: a $4 million round in 2017, a partnership with JetBlue, and eventually more than eighty paying clients, most of them praising the product even after it was gone.

The trouble was the size of those clients' budgets. A regional visitors bureau paying for an annual subscription is a real customer and a small one, and eighty of them does not fund a Seattle engineering team. By 2018 Berenstein was raising again, the round was not coming together, and he began talking to buyers instead.

~ two accounts of one collapse ~

A large travel company agreed to buy Utrip and went into diligence. What was supposed to take six weeks ran nearly four months. At the end of it there was a 94-page agreement ready for final approval and a close scheduled 48 hours out, waiting on sign-off from an oversight board inside the buying company. The board did not sign.

Berenstein told GeekWire he never learned why. In a later interview he gave a reason: the buyer's lawyers had found a patent held by a rival, WayBlazer, that conflicted with Utrip's technology, and that surfaced three days before signature. The two versions do not contradict each other so much as sit at different distances from the room, and both end with the same company having no money and no buyer in the same week.

~ the fallout ~

Gabrielle Andrews at the ATL Airport District Convention and Visitors Bureau in Georgia opened her email on the morning of 7 May 2019 and found a note from Berenstein: "It is with deep sadness that I inform you that Utrip, Inc. will be ceasing operations." Her organization had sent Utrip a $10,000 check for the coming year five days earlier, at Utrip's own prompting to renew. She called the bank. It was too late.

Other clients found out when GeekWire called them two months later. One tour operator in San Francisco was still inside a three-month contract billed in January and had heard nothing. utrip.com stopped resolving entirely and has not come back. There is no farewell page and no archive of the itineraries people built, just a domain that answers nothing on either scheme.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »