Am loving the recent run of interviews that I have been doing especially in the travel discovery and inspiration space. I have finally had a chance to write up my notes of an interview with Reuven Levitt of inspiration start up Tripbase. The interview and story came in part from my series of posts on the travel discovery and inspiration companies and part because Tripbase has just closed a round of series A funding, securing $2 million from private investors. I am going am going to do two posts on the interview. This first one will be on the company and the market. In part two I will share our discussion on start up fund raising.
What is Tripbase?
We have been talking on the BOOT about the different approaches that companies are choosing for generating recommendations. There is the criteria based selection process of Triporati, the event based selection of Joobili and the deal/price hunting approach of Voyij.
In Tripbase's case they have tried to distil trip recommendations critera to just five parameters - Nightlife, Dining, Shopping, Nature and Attractions. A search is then based on selecting your starting point (ie location), dates and (through using sliders) the relative positives and negatives of preferences the traveller has on each of the five parameters. For example I searched for a Spring (October) holiday from Sydney for a family of four with the following slider preferences hoping to reflect a "family" perspective (picture)
Search results are one thing but it is the refinement that is critical for recommendations. There is a need for a balance - enough to allow the consumer to find what they want but not too much to be overwhelming. The main refinement available from Tripbase are budget, weather and continent (Levitt is promising more).
Here are the results (picture)
I like the idea of the price refinement. That said, I moved the value slider a lot and did not get presented with a much greater variety of recommendations.
Where did they idea come from, where does the data come from?
Levitt and the rest of the Tripbase team came to the travel recommendation area not from a travel background but from a history of computational biology and building recommendation engines in the biochemistry field. This led him to attempt to build a generic recommendation engine and commence work on taking that engine and modifying it for travel. As he described it to me Tripbase is made up of two components rather than one. The first is the data centre and information collection mechanism. The second is the Tripbase website itself which sits above the data centre, providing and interface for customers to access the information.
They call the data centre the Travel Genome Project (not to be confused with Triporati's Vacation Genome Project). Levitt is claiming the TGP indexes and scrapes over 3,000 websites and therefore up to 22 million reviews. As Levitt put it "we use artificial intelligence techniques to make "sense" of a destination" drawn from the huge number of data sources being scraped and searched. So while Triporati is using technology and people/editors hand in hand to determine the results, Tripbase is relying on computational data analysis to sort through the overwhelming amount of data available - including the need to estimate prices and costs.
Thoughts on the market and monetisation
He has plans for both elements of the product and there are clients for both. At this stage the plans for Tripbase are further distribution and customer experience refinement (ie general competition in the discovery and inspiration space). For the Travel Genome Project plans, he was keeping his cards close to his chest, not giving away too much. If it is true that he has with this product a unique (or near unique) way of indexing destination information then he has a powerful tool for distribution through partners. But was not sharing a plan as to how he would take advantage.
Tripbase are not yet focused on monetisation and are in no rush. Levitt says he is not under revenue pressure yet and he wants to get the product right first. He contrasts this to the Big Four (Expedia, Orbitz, Travelocity and Priceline) who he saw as struggling to innovate and grow because they are, as Levitt put it, "struggling to focus on costs and spend marketing money while trying to cut back on fees and keep the business growing"
The basis of the product is an automated way of indexing enormous amounts of travel data through algorithms that link descriptive data, price data and sorting criteria. Very interesting. But there is still much to do before the product is a hit. While the search process is easy, I found that the refinement process did not produce the variety of options that I would have expected. The search process was much less complicated than Triporati but the results from Triporati were more creative, unexpected and intriguing. That said, it is hard to say which recommendation I would actually be more likely to act on- the creative unexpected one (Triporati) or the proven more anticipated on (Tripbase). Also I liked the the "cost per day" and average hotel and flight cost information provided with each recommendation. The price estimations seemed very accurate (as far as I could tell) and varied with time of year/season.
Monetisation will be a challenge but not an impossible one. As Levitt admitted the travel inspiration business is a "distance away" from purchasing behaviour and therefore will take time to target to advertisers.
In summary the product has made a great start. Some challenges to overcome but I am definitely seeing some a very entertaining and competitive battle starting to emerge between the different discovery and inspiration players. Do you have any thoughts on this space and the emerging battle lines?
Update - part 2 of interview here