Utrip, an itinerary-building service, nets $750,000 in funding
Some people don’t like to plan vacations in too much detail. For them, plotting an itinerary hour-by-hour feels like homework.
Yet others like to plan carefully, to avoid otherwise not allowing enough time to go between points A and B.
This latter group of well-prepared travelers may want to test drive Utrip, an itinerary-building Web app made by a Seattle startup of the same name.
Tell it your preferences, and it suggests detailed daily itineraries which aim to spare you from travel snafus, such as getting stranded in a neighborhood without a decent restaurant.
Today Utrip says it has raised $750,000 in a seed funding round.
Leading the round is Matthew Upchurch, chief executive officer of Virtuoso, a group of 335 luxury travel agencies. Also participating is CB Alliance, a New York equity investment group with an interest in big data plays.
The company has nine full-time employees, and it plans to use the funding to hire more people.
In the 20 months since Tnooz profiled Utrip, some things have changed.
A few weeks ago, the site came out of beta. It now allows users to do multi-city trips, up to 30 days in duration.
Until now, it has offered coverage for 27 European cities. But it plans to expand its coverage to a small number of major North American destinations soon. It will also launch a mobile app.
The basic premise of the app has stayed the same. In the company’s words:
Once a traveler indicates their interests and budget, Utrip’s algorithm sorts through millions of must-see sites, activities, events and restaurants recommendations to deliver personalized itineraries–in minutes.
There’s a lot of trade skepticism about the utility and money-making potential of trip-planning services.
That is partly because a raft of so-called trip-planning services launched around 2010 and were more aggressive in marketing than in engineering.
But Big Data analytics has improved with time. Sites like Netflix and Amazon have educated consumers about personalization. Travel is moving this way.
With the funding news, do you have any changes to your board?
Why are you any better than TripAdvisor?
TripAdvisor is a great signal, but we do things differently. TA generates a pre-determined list for everyone that ranks attractions from one to 100, such as a list of restaurants most highly rated by the site’s users.
We instead ask a traveler about his or her preferences via a broad questionnaire (such as by moving 16 levers and their ratings of past destinations). Then we leverage that information along with tags and the past actions of that user and similar users on our site to offer a more tailored recommendation.
I may figure out you like “Nature.” But does that mean that on this particular trip you want to go on a wildlife hike, a bungee jump, or visit a botanical garden?
I can triangulate the best recommendation by knowing about your stated and inferred interests in “nature”, “spontaneity”, and “adventure.”
We partnered with a psychologist at the University of Washington and spent 6 months trying to understand what the decision-making lifecycle is for a traveler to fine-tune our recommendations.
You say you’ve enhanced Utrip’s machine learning. What do you mean?
On the user side, we note the characteristics of a traveler and how he or she plans a trip in a particular city. When another user with similar interests plans a trip to the same city, we will adjust our suggestions according to what the first user liked.
On the destination side, we enhance the suggestions of our algorithm with human curation.
What do you mean by human curation?
We add human curation to the mix. We have experts on the ground — chefs, historians, photographers, microbrewery owners — give us advice on what’s worth seeing in their hometowns.
Is it a hurdle how much information you need to make accurate recommendations? Companies like Netflix and Amazon already have millions of records to draw upon for their recommendations, while Utrip is starting from scratch.
The average traveler spends less than a minute adjusting levers on interests and clicking labels for their interests and personalities. Then they can immediately go into our “day view.”
It’s not a 45-question survey. It’s a quick cartoony-and-cute interface. We get travelers into an itinerary as quickly as possible, and then we learn from their behavior.
Utrip does all of the work, while other sites give you an overwhelming universe of possibilities.
So, after the 16 levers are adjusted, Utrip goes out and generates 10,000 iterations of a relevant itinerary. It studies every previous iteration to get the best results in less than a third of a second.
We don’t recommend you visit a place on a day that it’s closed. We take seasonality into account, so don’t recommend an ice cream shop in winter. It’s much more customized than other sites.
One problem I’ve had with Utrip is that it sometimes generates itineraries that are unrealistic. For example, I plugged in London, and it claims I could go in 40 minutes from the National Gallery to Kenwood House. No way is that happening by public transportation plus walking through a large park. Ditto for the next suggested leg of getting from Kenwood House to the Household Cavalry in under 40 minutes.
We’re also continuously looking for reliable sources for transit times in destinations. The third-party data we rely on is variable in quality sometimes.
When we launch our smartphone application, things will improve. We’ll tap into the native mapping, to give more accurate estimates.
We also plan to add real-time traffic, so an itinerary will adjust on the fly. We’ll account for the hours of operation of a venue, if there’s a waterfall effect due to your delay.
Another problem I’ve had with Utrip is that sometimes I don’t want to book a hotel through it. It prompts me to see recommended lodgings regardless. I find that pushy.
A feature we’re working on is letting a traveler input a lodging that’s already been purchased, or to say if they’re staying with a relative and input that into Utrip.
What else is upcoming?
We’re working on strategic partnerships. We’d love to have an Airbnb relationship.
What’s the geographic breakdown of users?
Roughly 75% are coming from the US and Canada. About 20% are the UK and Australia. The rest is mostly Israel and Hong Kong.
About 70% of our visitors arrive organically through search.
Other companies have tried to tackle the custom itineraries market. For European itinerary-building, YourTour, a Belgian-based startup to help users create road trips, is the most direct rival. San Francisco-based GoPlanIt continues to target the US, but without much marketing mojo. Israel-based Plnnr appears to be a ghost town. Why do you think you’ll triumph?
Of those, I’m only interested in GoPlanIt as a serious competitor. Our intelligence offers significantly more personalization. If you go through its experience, you’ll see the limitation of itineraries based on limited information.
Overall their product is wonderful. But we feel ours clearly stands out as superior. For us it’s about complete personalization of every aspect of trip-planning.
What’s your revenue model?
Our business has an affiliate model. The tool works with multiple inventory providers to let users optionally book some or all of a trip in one click.
As we scale, we can monetize more aspects, such as travel insurance and cell phone plans.
Is your company bucking the trends?
We’re riding a growing wave.
Our approach is now validated by industry trends. PhoCusWright’s Travel Technology report talks about several trend drivers, including the importance of content curation and Big Data in helping travelers plan their trips.
RELATED, FROM TNOOZ
Sean O’Neill is the Editor-in-Chief of Tnooz and is based in southern New Jersey, in the US.
Before joining us, Sean was a regular contributor to BBC Travel, a senior editor of BudgetTravel.com, and an associate editor at Kiplinger’s magazine.
Follow him on Twitter (@sean_oneill).