UberX and UberEats: Setting the Take Rate | In Practise

UberX and UberEats: Setting the Take Rate

Former Global Head of Pricing and Strategic Initiatives at Uber

Learning outcomes

  • Core drivers of UberX and UberEATS pricing strategy
  • How to use price to balance supply and demand in marketplace businesses
  • How economies of scale will drive Uber’s ride-hailing take rate
  • Potential pricing strategies for Lyft versus Uber
Print

Executive Bio

Kapil Agrawal

Former Global Head of Pricing and Strategic Initiatives at Uber

Kapil joined Uber as the Global Head of Pricing & Strategic Initiatives in 2015 when the business was rapidly growing across developed but also emerging markets. He was responsible for formalising pricing structures across both UberX, UberPOOL and UberEats globally. Kapil left Uber after two years and joined Poshmark, the fast growing social commerce marketplace, as VP of Finance and Corporate Development where he has helped quadruple revenue at the business. Kapil has led $90m of capital raising for growth equity companies and has deep experience scaling marketplace businesses. Read more

View Profile Page

I think a good place to start would be if we could take a step back to when you joined Uber. How were you thinking about pricing for the ride-hailing business?

I joined Uber in 2015, when it was expanding, very rapidly, in some of the emerging markets, like the Asian and the developing markets. At the time, they were also developing new products, like Uber Eats and Uber Pool. Our top process for Uber’s pricing, especially for ride-sharing, was that there were a few areas we wanted to focus on.

First, we wanted to be the cheapest ride-sharing player in the market and that would mean two things. Firstly, we would like to be at least 35% to 40% cheaper than a taxi. The way we thought about this product was, we are not only competing with the taxi, as a product, but we also wanted to compete with public transportation or personal vehicle ownership. If you keep the price of your product, 35% to 40% cheaper than a taxi you could, potentially, compete with public transportation and, also, personal vehicle ownership. That was one thought process that was in our minds, when we were going to a particular market.

We also wanted to make sure that we have back-calculated pricing in such a way that drivers are making enough money, in that particular market. The way we thought about what was enough money, was to look into what were the other top employers in that market. For example, if you think about the US market, you will have Walmart or Starbucks, who are the largest employers. Let’s say, they are paying $15 to $16 per hour, we would like to make sure that, in the long term, on the platform, if the drivers achieve X percentage of efficiency, they are able to make that amount of money, or more than that.

The way we think about efficiency is, the amount of the time that the drivers have the app switched on, that would be your numerator. The denominator would be, for how much time they are on a trip. For each city, we will think what your long-term efficiency, for that market, would look like and, for that efficiency level, we’ll set pricing in such a way that Uber’s drivers are able to make enough money. By taking those two factors into account, one, on the rider side, we want to make sure that the product is 35% to 40% cheaper and, on the driver side, that they are able to make enough money, at a certain efficiency level.

To reach that particular efficiency level, it may take a certain amount of time; it may take six to nine months. Whilst you are working to reach that efficiency level, you want to provide a driver subsidy and a rider subsidy. The rider subsidy is used to make sure that you have the network effect kick in. The driver subsidy is because, during that time that they are reaching that efficiency, they still need to make enough money in the market. That’s how, on the overall picture, we think about the pricing.

Driver’s earnings are actually a core focus? You would back-out the utilization from that $15 an hour, for example, that you want the driver to earn?

Yes. We had a lot of data, from launching in many cities, so we knew what the potential efficiency would look like, in a particular city. We had the characteristics of different cities. For example, San Francisco, LA, New York, they can combine into a particular group of cities and we can assume that, yes, those cities could reach 60% to 70% efficiency, whilst the other set of cities, like Dallas, Houston, etc., they could, potentially, be 50% to 60% efficiency. When we are launching in a new market, we could say that, okay, this market could resemble this city and what the long-term efficiency for that particular market would be.

If we achieve that efficiency, in the six to nine months, what should the driver pricing be so at that efficiency level, the driver can still make sufficient money, in that market.

This is why scale is so important, because the more drivers you have, the more riders, the higher the frequency, the lower the cost. This makes it cheaper for the rider and more rides, means more earnings for drivers?

Yes, that’s correct. The cheaper the product, more people will use it. As you have more and more people using it, that specifically means that the driver opportunity will go up. Let’s say there are more and more people using it, there will be one ride after another, so there won’t be any wait time for the drivers. As their efficiency increases, they will be able to make a similar amount of money, even at the lower prices for the riders.

For example, let’s say we started at a $10 ride, for the rider. At that time, the efficiency, because of the level of demand, was 40%. If you cut the pricing to $5, in that particular market, the demand will go up, exponentially. What would that mean? The driver’s efficiency could increase from 40% to 60% or 70% and they can still make the same amount of money, as they were doing when the prices were $10 for the rider.

Although, in some of the markets, if you don’t see that a certain level of efficiency kicks in, or a certain amount of ride volume go up, as we cut the prices, we are open to roll-back those price cuts that we used to do.

What examples do you have of where that didn’t work?

For example, we did do this price cut, at the time, in hundreds of cities in the US and other parts of the world. Out of all those cities, we’ll track the specific metrics and the driver earning and efficiency levels were two of the metrics. We tracked those metrics, to enable us to understand the health of those markets, whether there was enough of a balance between the supply and demand, after the price cut in that market.

Two things could happen after the price cut. Either your demand goes up, significantly and that means that the driver’s efficiency goes up and drivers are making enough money, even at the lower prices for the rider. The second situation could be, you cut the prices but demand does not go up, significantly. Because of that, drivers are not making enough money and drivers churn on the platform and then you will see higher surge and smaller number of drivers to receive, and higher ETAs. If you see those types of metrics, then you must be open to consider rolling back the price cut, so that the drivers can make the same amount of money that they were making earlier.

How do you balance that supply and demand? Is efficiency the key metric here that you need to focus on, in all cities?

Not really. Efficiency is, definitely, one of the metrics that we consider. But there are a set of several metrics that are very important for us to look at, to understand the supply and demand of that market. We’re looking at both supply-side metrics and demand-side metrics. On the demand side, we’re looking at surge, so how many trips are surging, as compared to requested trips. How many people are requesting and, out of those requests, how many trips are being completed. Those are mainly surge and complete to receive, on the demand side.

On the supply side, efficiency is definitely one of the metrics. ETAs, estimated time of arrival, was another metric. Those were a key set of four metrics that we will look at, in any of the markets, to understand whether it is a supply-constrained market or a demand-constrained market. For example, within a particular city, they have higher surge, lower C2R. Higher surge will indicate, yes, it’s a supply-constrained market, so that’s why you have a lot more demand and lower supply. If there is a higher C2R, it means that a lot of trips are being completed, which would mean that it is a demand-constrained market. Lower efficiency will indicate, if it’s a demand-constrained market, you can, potentially, increase the demand to increase the efficiency levels, or it’s an over-supplied market. The fourth one, as we talked about, if there are higher ETAs, it would mean that there is not enough supply. If there are lower ETAs, you have a lot of supply.

Based on those, we’ll think about what is the best action to take, for that particular market. We look at those four different metrics and, depending on higher or lower, for each of those metrics, we can come up with different strategies for that specific market.

Sign up to read the full interview and hundreds more. No credit card details required.
Sign up to read

Audio

UberX and UberEats: Setting the Take Rate

February 12, 2020

00:00
00:00
Sign up to listen to the full interview and hundreds more. No credit card details required.
Sign up to listen