We Are Experimenting With Speech Recognition and Other AI Techniques: OYO

In an exclusive interaction with Dataquest,  Anurag Gaggar, VP – Product, OYO  talks about the OYO app, the IT initiative that impacted the company’s business, Data Analytics and AI. Excerpts:

The gradual progression of OYO has been app-driven. What changes have come in the app itself and the entire ecosystem of consumers and hotel partners?

We launched our app in May 2015, the first version allowed people to make bookings with some details about the property and basic search features on maps. The first version of the app took about 2 months to develop. However, all products keep improving iteratively and more work keeps happening all the time to refine the experience for users. Since then, there has been a lot of refinement. As the app size or user base increases, we have to make sure that their needs are well met. There has been 10mn app downloads till date across Android and iOS platforms. The app contributes 80%+ of our transactions and 60%+ of our traffic.

We started with limited number of hotels in cities. For example, if a person asked for all the options available in Gurugram, the app would show 20 of them to pick one. As the network grew, more people started browsing. We realized that even in Gurugram, we might have 300 to 400 options, but not everyone will be interested in browsing through all of them. So the user interface started asking for more preferences to narrow down the search options.

We improved features constantly and forged more partnerships. Online payments were introduced. We now have partnerships with Paytm and Mobikwik for seamless quality integration which allows cashless transactions. We launched Oyo Captains with a team of over 250 people on the ground to resolve customer issues. When a customer makes a booking or checks into a hotel, a captain is assigned to resolve all queries instantly.

OYO has features that connect you to the hotel’s Wi-Fi automatically without asking for a password. Customers have an option to order food in certain hotels which we manage ourselves. In Townhouse and flagship hotels, we have tight control on these options.

How do you place yourself against competition? Goibibo and Makemytrip have penetrated the market at a much faster pace than OYO over the past two years. What sort of satisfaction have you achieved with hoteliers and customers?

Makemytrip and Goibibo offer are pure-play common travel agencies that helps in selling two or more flight tickets or bus tickets. We are in a different business model and segment that we operate in. So we think of ourselves more as a hotel chain or hotel provider who is not only just doing the job of selling a room to a customer through the app, or through whichever channel, but also ensuring that hotels which come on the platform are standardized to some degree.

A customer knows that the end-to-end experience will be predictable, consistent and good. A lot of hard work goes in building a technology to audit these hotels and track their overall performance. It is not just providing travellers how to get into a hotel but also how do they check in and out. Can I give the room number of their check-in? Can I enable them to order food? Can I connect them to the Wi-Fi?

A bunch of these things are possible because our relationship with these hotels is strong. We have technology deployed at the hotels to run those operations. The core thing users get from OYO is the holistic experience and they find it really useful which starts from the search-and-booking process. Having a customer-friendly app is very important. It does not end with search-and-booking but goes all the way through to the point of checking out.

Is there any IT initiative that has impacted the company’s business the most recently?

The launch of OYO Money had a significant impact on the business leading to exponential growth during the initial months of the app launch. It is a strong driver for us in terms of new user acquisition and transactions. We have also worked on personalizing the app for our users and it has shown a positive impact on conversions. It is also a meaningful way for us to offer better customer experience.

What are the key areas and functions where you are deploying data analytics and AI?

Millions of customers visit our online properties across our consumer Apps, oyorooms.com and mobile web site each month to discover and book their favourite OYOs and manage their post booking experience. All these interactions along with guest interactions at our properties and other channels like call centre generate a huge volume of data every day. We apply advanced data analytics and machine learning to learn from this data and use this to further refine guest experience by providing them highly personalised recommendations. We also use this data to identify any potential risks to guest experience by detecting issues at our hotels and proactively mitigating these.

We are experimenting with speech recognition, natural language processing, deep learning and other AI techniques to improve the guest experience at our hotels. These along with IoT have the potential to transform the hospitality experience of the future and deliver an unparalleled convenience to guests by enabling automated check-in and check-out, controlling room lighting and AC temperature and enabling food orders and other service requests using natural speech.

Please cite two examples of how usage of data analytics and AI has helped you in improving customer service levels?

Data analytics has enabled us to provide our guests a highly personalized and convenient experience throughout their interactions with OYO. When you are searching for your next OYO, the search results including sort order is highly personalized for each guest, so they can find the best hotel for their needs with minimal effort.

We provide our guests the flexibility to pay online or at the hotel. Pay at the Hotel guests tends to have a high cancellation or No-show rate and some of these bookings are not authentic. During peak periods many of our popular hotels often get booked at capacity well in advance resulting in many guests not being able to book these. We use data analytics to identify and block potentially unauthentic Pay at Hotel bookings, thus by protecting the ability for genuine guests to avail these hotels.