Before we dive deep into Uber’s business model and the role of data analytics in optimizing the business, let’s start with some gentle introductory discussion to understand how Uber works better.
Uber is a digital aggregator application platform, connecting passengers who need a ride from one place to another with drivers that are willing to serve them. Riders create the demand; drivers supply the demand and Uber acts as the facilitator to make this happen seamlessly on a mobile platform through its engineering.
Following canvas lays out the key components of Uber’s business model in a nutshell:
Since Riders and Drivers are the most essential part for Uber’s business model. Important value proposition for riders which Uber offers are:
- On-demand bookings
- Real-time movement tracking
- Accurate Estimated Time of Arrival (ETAs)
- Cashless Payment
- Lower wait time
- Upfront ride fares
- Multiple cab options
Similarly, Uber’s value propositions for drivers are:
- Flexibility to drive on their own terms
- Better income
- Lower idle time to get new rides
- Better trip allocation
This brings us to the question about how Uber makes money or what all revenue stream Uber has? From a high level, Uber takes commission from drivers also called partners for each ride but this is not the only revenue stream for Uber as it uses different methodologies to increase revenue some of them are:
- Commission from rides
- Premium Rides
- Surge Pricing
- Cancellation Fees
- Leasing Cars to Drivers
- Uber Eats and Uber Freights
Role of Analytics and Business intelligence in Optimization:
Uber has a massive database of drivers, so as soon as you request a car, Uber’s algorithm goes right to work and it matches you with the driver closest to you. In the background Uber is storing data for every trip taken — even when the driver has no passengers. This data is leveraged to predict supply and demand, as well as setting fares. Uber also looks at how transportation is handled across cities and tries to adjust for bottlenecks and other common issues.
Uber also gathers data on its drivers. In addition to collecting non-identifiable information about their vehicle and their location, Uber also monitors their speed and acceleration, and checks to see if they are working for a competing company as well (such as Lyft).
All this data is collected, crunched, analyzed and used to predict everything from the customer’s wait time, to recommending where drivers should place themselves via heatmap in order to take advantage of the best fares and most passengers. All these items are implemented in real-time for both drivers and passengers alike.
Uber’s biggest uses of data comes in the form of surge pricing, a model nicknamed “Geosurge”. Uber leverages predictive modeling in real-time based on traffic patterns, supply and demand.
In the short term, surge pricing substantially affects the rate of demand, while long-term use could be the key to retaining or losing customers. Customer backlash on rate-hiking is strong, so Uber has considered using machine-learning algorithms to predict where demand will be strong, so that drivers can adequately prepare to meet that demand, and surge pricing will be significantly reduced.
Keep in mind, however, that 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. And if you think all major metropolitan cities are alike – think again. Just look at how New York City compares to London:
Of course, collecting all this information is just one step in the big data journey. The real question is — how does Uber determine the best way to make decisions using this information? How do they glean actionable points out of the data they collect? For example, Uber manages billions of GPS locations. Every minute, their platform juggles millions of events. How do they leverage these details into a way to better manage moving people and things from place to place?
Their answer is data visualization.
Data visualization specialists range from computer graphics professionals to information design. They handle everything from mapping and framework developments to data that the public see. And a lot of these data extrapolations and visualizations have never been done before, which has created a need for tools to be developed in-house.
Without getting too technical, some of the many applications for their data visualization challenges include:
Mapping Applications for City Ops Teams
But these aren’t just data visualizations for engineers and data scientists to pore over. Data visualization also helps the public better understand what Uber does and how it works, such as this visualization of how uberPOOL helps reduce traffic
Another example is of particular importance in big cities, where understanding the density of a given area may lead to dynamic pricing changes. Uber demonstrates this with a combination of layers that let them drill-down to see specific areas in more detail:
Apart from Data Visualization, Forecasting is an important Business Intelligence technique Uber uses to optimize future processes.
Marketplace forecasting: A critical element of the platform, marketplace forecasting enables Uber to predict user supply and demand in a spatio-temporal fine granular fashion to direct driver-partners to high demand areas before they arise, thereby increasing their trip count and earnings. Spatio-temporal forecasts is still an open research area.
Learnings from Data:
It’s one thing to demonstrate how Uber uses data science, but another completely to discover what their findings mean beyond just a ride on-demand. Uber teaches us a great deal about using big data – and not just sitting on it. They also teach us to look for connections in every possible ounce of that data. Every time you’re collecting information but not using it, there’s the potential for a missed opportunity to grow and improve your business.
It’s also worth realizing that, much in the same way as Uber records the pulse of a given city, that not every answer you glean from your data can be carried over and applied in a blanket-like fashion to a completely different city. That gathering the data independently and analyzing it for what it is and what’s going to make the insights and opportunities pop out.