Thanks to the app based car booking service, users who need to get somewhere is able to connect with drivers willing to give them a ride just by a click or two on their smartphone. This service started in San Francisco, then Uber quickly became a huge hit and new cities and countries were added not long after that, including Indonesia.
Uber’s huge hit is actually not surprising due to its simplicity. In order to get a ride from an Uber driver, all you have to do is simply open the app, set the pickup location, request a car, get picked up, and pay. But behind this simplicity, there’s a great deal of data analytics going on to make all of this happen in a smooth process.
Yes, Uber is rooted firmly in big data and utilizing this data in a very effective way. While Uber moves people around the world without owning any cars, data moves Uber. From surge pricing, driver ratings, to estimating fares, Uber heavily relies on the big data to build an intelligent company. The more data they collect, the more information they can derive from patterns and behaviors, thus allowing them to continually improve their service.
A huge database of drivers and users
The first thing you should know about Uber’s behind the scene process is that the company has a huge database of drivers. Even when these drivers are not carrying any passengers, they actually still continue to generate data for Uber. It’s because they transmit data back to the central platform at Uber, which is used to draw inferences about traffic patterns. Uber will store this data in the database for supply and demand algorithm analysis, as well as for autonomous car research, surge pricing, monitoring driver’s speed, and tracking the location of drivers.
On the other hand, as soon as you request a car, Uber’s algorithm matches you with the driver closest to you. Behind the scene, Uber is storing data for every trip user has taken. This data is leveraged to predict the demand for cars, set the fares, and allocate sufficient resources.
What’s more interesting is that Uber’s data science team also performs an in-depth analysis of the public transport networks across different cities. This enables Uber to focus on cities that have poor transportation and make the best use of the data in order to enhance customer service experience.
Bridging the gap between demand and supply
Uber has this one system called Geosurge, or as users call it: surge pricing. It works like this: say you’re running late for an appointment and you need to book a ride in a crowded downtown space. When you order an Uber, you’ll most likely have to pay almost twice as much for it.
It turns out that the Geosurge is one of Uber’s biggest uses of data. The system is actually similar to the pricing strategy used by hotels and flights for their holiday fares and rates, except Uber implements a predictive modeling in real-time based on traffic patterns, supply, and demand.
You might think that this Geosurge system would make people hesitate to order an Uber at a certain time. Yes, you’re right, because we’ve also been there. But, here’s the thing, Uber doesn’t just activate the Geosurge system anytime they want. They utilize the data science to analyze the short-term and long-term effects of surge pricing on customers. In the short term, surge pricing affects the rate of demand. Meanwhile, the long-term effect could potentially be the key to retaining or losing customers.
In order to bridge this gap, Uber uses machine-learning algorithms to predict where demand will be strong. This will allow drivers to prepare to meet that demand, and surge pricing will be significantly reduced. According to Kissmetrics, Uber does this by building an in-house data wrangling system, which monitors millions of system interactions and metrics, as well as alerts engineers in cases of serious outages.
However, the supply and demand data are not the same from city to city, so Uber engineers devised a way to map the “pulse” of a city to connect drivers and riders more efficiently.
Implemented in real-time
What makes Uber become a huge hit is that the data science-driven insights don’t just stay within the dashboards or company reports, but they are implemented in real-time. The aim is, of course, to create a positive experience for customers and drivers.
Geosurge, for example, is based on both geo-location and demand for a ride to position drivers efficiently—just like explained above. Besides Geosurge, Uber still has other data products that are also implemented in real-time.
When you order an Uber and before you click the “request” button on the app, Uber shows you an estimated fare you will have to pay for the ride. They calculate fares automatically using street traffic data, GPS data, and their own algorithms that make alterations based on the time of the journey. That doesn’t stop there. Uber also analyses external data like public transport routes to plan various services.
Every time you use an Uber service, you will be asked to rate the drivers. On the other hand, the drivers can also rate you. Uber relies on a detailed rating system to build up trust and allow both parties to make informed decisions about whom they want to share a car with. In particular, this rating system will help drivers to maintain their standards high.
For drivers, they also have another metric to keep in mind, which is acceptance rate. This refers to the number of orders they accept versus the ones they decline. Uber wants their service to be consistently available to passengers, so they expect the drivers to keep the acceptance rate above 80%.
With a neat, reliable big data analysis, it’s no wonder that Uber has become one of the fastest growing startup standing at the top of its game. Today, Uber has more than 8 million services, 1 billion Uber trips, and 160,000 people driving for Uber across 449 cities in 66 countries. Uber is known as a remarkable name to reckon when it comes to solving problems for people in transportation.