The travel industry in India is buzzing with startups and established players innovating quickly to get ahead of the pack. Portals like MakeMyTrip excel in airline and hotel bookings. Others focus on destination discovery, providing curated content. Now, big data and machine learning are creating new possibilities.
In a previous article, we saw how Bangalore startup FindmyCarrots uses natural language processing (NLP) to help people find destinations or hotels. For six years, the US-based Hopper also tried to do that with different tools before pivoting to predictive analytics of flight prices. The Hopper homepage today offers to let travelers choose “when to fly and buy”.
Arijit Mukherjee, co-founder of FindmyCarrots (FmC), says these moves validate his premise that travelers are game for interesting tech and are demanding better solutions from the online travel industry. He believes FindmyCarrots has built up the semantic capability to outdo other players in this space.
Our differentiation is already in front of you. If you run some ‘conversational’ searches, you can see the results we produce vis-a-vis those of other travel search engines. And with us improving the results every passing day, the future can only be more fun and informational for our users.
Taming the information overload
A survey by travel portal Yatra last month found that three out of five travelers in India go online for booking and itinerary planning, but a fifth of them are overwhelmed by the information overload on the Internet and end up going to traditional travel agents. A related finding was that most travelers now look for meaningful and off-the-beaten-track experiences. A semantic search engine could address these pain points.
On FindmyCarrots, if you search for a “beach within 200 km of Bangalore”, you don’t get a google of options — because in reality there are no beaches in that area. So you only discover beaches as you widen the circle on an interactive map. At around 270 km, you hit the beaches near the former French colony of Auroville in Puducherry on the east coast, and Meenkunnu beach in Kerala on the other side.
“We have to analyze a humongous amount of data to be able to produce this gradually increasing search based on distance information from a place in question, which in this case was Bangalore,” explains Arijit.
A tool for discretionary travelers
FindmyCarrots appeals to discretionary travelers who use the internet to do their own research and find novel experiences. A PhoCusWright report says discretionary travelers account for one-tenth of the US$10 billion Indian online travel market, which resulted in a compounded annual growth rate of 30.6 percent from 2011. More than two-thirds of consumer ecommerce transactions in India relate to online travel currently.
But FindmyCarrots only views the Indian travel market as a launchpad. It had global ambitions right from day one. “We use machine learning and do not curate our content. The technologies we use are linearly scalable and highly available. So, we have all the building blocks needed to scale massively and rapidly,” Arijit says.
The startup has seen modest growth since its launching in September 2013 — from 50 visitors a day to daily traffic of over 1,000 now. “What has been a more inspiring number for us is the average time spent on FmC by our users. It has consistently been above three minutes. Now, with newer features like hotel insight and booking options, our users are spending a little over five minutes on an average. We also have had significant traffic from outside India and they too spend similar time on FmC to find Indian destinations. This makes us confident that we can make this Semantic Travel Search Engine a global phenomenon,” says Arijit.
The vacation that started it all
The idea came about when Arijit was planning a vacation in the summer of 2012. “We wanted to go to a hill station in north east India. When we started searching for it, we either got dated and irrelevant information or we were overwhelmed with data. Finally, we gave up looking for an offbeat destination and headed to well-known Shillong. On our return, I decided to see if I could build something that would answer such travel queries more intelligently,” he says. That quest consumed all his nights for almost a year when he held a full-time job.
Arijit holds a Bachelors in Engineering from Jadavpur University, majoring in Computer Science, and an MS in Software Systems from BITS, Pilani. In his professional career before starting FindmyCarrots, he worked with Nortel Networks and Polaris Wireless. His friends, Sivaramakrishnan Nageswaran – again a BITS, Pilani grad who worked with Polaris Wireless – and Kanwar Bir Singh Sangha – an engineer who helped build a multi-million dollar product line for Mavenir Systems — joined hands with Arijit to start the company. They bootstrapped, built and launched the beta version a little over a year after it was conceived.
Leveraging AWS, Google Analytics and Clicky
Like all bootstrap heroes, FindmyCarrots too had to manage on a shoe-string budget. The team made do with free tools available on the cloud wherever possible. Initially, they hosted the site on a free tier of Amazon Web Services (AWS) and then upgraded to a higher-end AWS when the increased traffic demanded an upgrade. They used free versions of web analytics tools like Google Analytics and Clicky to understand their strengths and weaknesses, before paying up for a complete version of Clicky. Now, as the startup’s technology gets validation and there are more takers for it, new possibilities are opening up.
One scheme is to rent out the data analytics and semantic technology that FindmyCarrots brings to travel. “We can white-label our platform to bring its power to big online travel agencies who want to differentiate themselves from their competitors and provide a solution that travelers have been demanding for quite some time now,” says Arijit.
The semantics of online travel appears set for a sea change. There’s more on that in the first part of this article.
(Top image: Flickr user Mehul Antani)Editing by Terence Lee