Ixigo predicts machine learning will make the difference

Ixigo.com has claimed the number two spot in India’s online travel market and believes that its growth comes from using machine learning techniques to provide products which solve the problems of real Indian travellers.

Its CEO Aloke Bajpai told Tnooz that Ixigo is now second only to the combined MakeMyTrip/Ibibo Group in terms of reach. Using independent third parties such as AppAnnie and Kalagato,  the latest stats show that Ixigo’s three apps – rail, flight and hotels – have been downloaded 30 million times, with the monthly active users coming in at 10 million and daily active users at one million.

He said:

“We see ourselves as a product and tech company as much as a travel business. We have just over 100 people, 70% of whom are engineers, and that’s different from our peers. We are agile in terms of our tech, our approach to the customer experience and we are pro-active when it comes to solving problems travellers face.”

Its rail app is the most used in India and typifies Ixigo’s tech-driven approach to problem solving. One idiosyncracy of the Indian rail system is that passengers can reserve standby – known as wait-listed –  tickets and get a PNR notification from the train operator if the seat becomes available for travel.

Ixigo recognised this as a problem for many travellers and built a “PNR predictor” into its rail app, which uses big data and machine learning to work out the probability of a standby booking becoming confirmed. Its rail app is now analysing three million PNRs a month and is becoming more accurate the more it learns.

Ixigo’s rail app does not sell tickets, but Bajpai said that it was “exploring ways” with the IRTC to add this feature.

For now, the app operates as an information portal for Indian rail users. The recent addition of a “social layer” for users to rate trains, stations and facilities is positioning the app as “the TripAdvisor for Indian trains.”

Ixigo is monetizing the traffic to the app by upselling and cross-selling various products, such as taxi rides from Ola and Uber, which are accessible via Ixigo’s cabs app and integrated into the rail app. It also upsells hotels to rail travellers and there are a small subset of rail travellers who are in the market to be upsold a flight if their train standby is not confirmed.

Rail travellers represent the “heart of India” with Bajpai noting that a lot of the traffic and interest in rail comes from commuters rather that tourists, which is another difference between it and its peers.

However, Ixigo is benefiting from two macro trends in India which are driving the growth in domestic  travel, and by extension international travel as well.

“It’s not just that there is a 25% year-on-year increase in new air capacity coming into the market, it’s that the fares are more affordable to more people.

And for hotels, years of discounting by the big OTAs has told consumers that hotels can be booked online at a great rate.”

Machine learning and predictive algorithms have also been at the heart of its flights app, feeding into a “fare predictor”, a fairly common-place flight meta feature. But Bajpai believes that the Ixigo has taken this one step further and by concentrating on the top 20 city pairs in India it can also predict when an airline or rival OTA is about to have a sale, “so we are looking at predicting promotional activity as well as the actual fares.”

Ixigo will extend this to the top 100 city fares and also work on a  fare predictor for international routes.

One characteristic of the Indian online travel market has been the drive for market share and growth – MakeMyTrip talked about growing market share in the short-term at the expense of profitability even before it received the Ctrip investment.

Discounting and marketing are two mechanisms used to grow share in India. On discounts, Bajpai said that most of the discounts offered by Ixigo were co-funded by their partners, limiting its exposure.

In terms of marketing, Bajpai said that Ixigo is not signing up Bollywood superstars as its brand ambassador. He said that Ixigo is getting more efficient at non-paid search and content marketing and is seeing engagement levels comparable with TV campaigns on its owned social media channels.

“We’re always looking at the path to profitability and we’re not too far away,” he said. “Our EBITDA margin is getting better as we grow our revenues and keep our cash burn rate under control.”


Source link

Leave a Reply

Your email address will not be published. Required fields are marked *