Arvoia
Nominated Award: Best Use of AI in a Consumer /Customer Service Application
Website of Company: https://www.arvoia.com/
Arvoia offers its clients in the travel and mobility sector sophisticated, readily deployable Ai driven prediction and personalisation products. Arvoia enables its clients in the travel and mobility sector to gain deep direct customer understanding by deploying our prediction and personalisation products driving revenue uplift and ultimately a better customer experience. The Arvoia platform enables our customers to train and deploy predictive model in days and see increased performance within weeks. Arvoia have worked in travel and mobility for over 25 years and during that time have implemented solutions across the entire travel spectrum, including car rental companies, airlines, brokers, hotels, OTAs, GDS and more.
To date, Arvoia have built recommendation strategies, solved social mobility challenges, expanded and enriched the content offering of companies throughout travel and mobility. Through all of this our relentless focus has been on dramatically improving customer experience and satisfaction, whilst driven double-digit revenue growth. We have done this by identifying, analysing and solving customer problems which are uniquely specific to travel, through the creation of highly targeted machine learning solutions.
Reason for Nomination:
The average person visits 38 sites before booking a trip or vacation, a long, laborious and unenjoyable task. The travel industry is changing, and now adays we are our own travel agents. One sector of the travel industry which has seen huge advances in choice in the last number of years is the mobility industry. The customer has never had more choice when it comes to transportation with private transfers and ride share apps like Uber, competing with the car hire industry.
Car hire is renowned for its poor customer experience, with the user flooded with available cars and prices. This results in users trying to negotiate lists of cars which are described using confusing industry standard terminology and a sense of apprehension with respect to price and hidden fees. If the car hire industry wants to compete in the modern-day sector, they need to address these issues to enable the customer to find the car they want quickly and easily. Arvoia enables the client to do this through our sophisticated, readily deployable Ai driven prediction and personalisation products.
When someone visits a car hire website, they input the search details (i.e. location and date ranges) and are shown the list of available vehicles. However, the cars are displayed from low to high and use industry specific terms which are confusing, so the user often finds it hard to find the specific car they want based on their needs and price point. On average, only 4% of searches made, result in a booking, and almost 70% of people who visit a car hire site make one search but then leave the site and never return. We have developed the Arvoia Ai Framework which helps to overcome these issues by using Ai solutions to show the right car to the right person at the right time and thereby improve the overall experience.
Our Ai framework uses several data science techniques, including supervised and unsupervised learning methods. We use supervised methods like neural networks, ensemble models, decision trees, etc to predict which car suits the customers’ needs at that time and unsupervised methods like nearest neighbours and methods like DBScan to identify meaningful clusters of customers based on their behaviours and to identify trends in demand and price in an industry which is massively impacted by seasonality. We also take advantage of several technologies and tools to ensure our framework is robust and scalable, these include Knime, R, Python, H2O, Weka, PMML, Amazon Web Services S3, Redshift and many more.
Within weeks of Arvoia’s personalisation products being used on the client website, we see significant improvement in the user experience as well as increased bookings and revenues for the car hire company. Looking at the impact on customer experience we see 4 key results:
1.Increased Click Through Rates: Users who see an Ai ordered result are more likely to click on a car in the search results to proceed to the next step in the booking flow compared to users in the control group. This indicates the user is more engaged in the booking experience.
2.Reduced searches: The average person makes 3 searches before making a booking. The users who saw Ai ordered results made on average 2 searches before making a booking, meaning the user doesn’t need to spend a long time moving between sit es and making multiple searches.
3.Reduced scrolling: Users in the test group booked the first car displayed 40% of the time, but 75% of the remaining users booked a car within 3 slots of the predicted car, reducing the need to scroll down the page.
4.Reduced sessions: Users in the test group who made a booking had an average lead time (time between date booking and rental date) of 7 days, whereas the test group had average leads time of 10 days. This means users in the test group are not returning to the site to do additional searches across multiple days before making a booking
As a direct result of improved customer experience, Arvoia ’s typical client sees a 20-25% revenue uplift, calculated using a 20% control group. This is made up of 2 parts:
1.Increased bookings : More users completing a booking resulting in an average 4-5% uplift over control group
2.Increased revenues: More relevant (usually more expensive) cars being booked by users resulting in an average 15-20% increase in revenue