Uber Experience: The Tale of a Dynamic Marketplace

As I ruminate on my year long stint at Uber, I can’t help but miss the dynamic work that I performed every day. One thing about Uber: it’s on 24x7, 365 days until once in a century pandemic hits. And it hit Uber hard. Really hard. So much so that 30% of its dynamic, brilliant workforce had to be laid off and are scrambled to find work during the pandemic. I am one of them.

However, that’s not the objective of the post. I am really grateful for the incredible opportunity, and as I reflect I have nothing but the great memories of the past so much so that I decided to delineate the nature of work in the Uber world -- what makes it dynamic, interesting, puzzling, and fun.

I was part of the Marketplace team, the heart of everything that Uber does. Uber’s marketplace denotes the prime example of the demand-supply dynamics from the Economics 101. For that reason, you would see economists love to cite the Uber case to explain many of the demand and supply phenomena. It reflects the market’s power to self-organise. Adam Smith would be proud had he was alive to see how the riders and drivers self-organise to make the transactions happen.

At its core, Uber serves as the matchmaker. A rider needs a ride to get to somewhere he or she needs to. A driver on the road roaming freely wants to give a ride. Uber matches them to make the transaction happen. Uber takes a fee from the value of the transaction. The transaction fee depends on the market it operates, ranging generally from 15% to 25%.

That means Uber needs to make each transaction happen as fast as possible. Every time that a rider does not get his request honoured in his platform is a loss for Uber. And every minute that a driver does not carry a rider or passenger on a paid trip is earning loss for both the driver and Uber. Here comes the role of a marketplace manager -- my role.

A marketplace manager liquidates the Uber market. If there are more riders than drivers in a given area and a given hour, the role of the marketplace manager to bring in more drivers in that area at the same time to match the demand i.e. matching supply with demand. For that, the marketplace manager constantly looks at the marketplace metrics and optimises them.

The holy grail of the marketplace metrics is C/R, meaning completed/request ratio, the fulfilment ratio. For every request that comes in, how many of them Uber are able to convert into completed trips. Optimising for C/R means optimising for the marketplace. If C/R is falling because demand is outstripping the supply faster, then the marketplace manager needs to think how to bring in more supply. If C/R is too high, that means supply is stable and absorbs more demand or demand might be drying up, the marketplace manager might need to work on increasing demand. And the marketplace manager has this superpower called "incentives" to iron out the inefficiencies of the market. Often, the Uber's cut -- the 15% to 25% fee that it takes as the matchmaking fee -- are utilised to incentivise the riders to request more trips and drivers to drive more.

This illustrates a minuscule portion of the story of managing the marketplace. Under the hood, there are hundreds of other metrics at play depending on the different geographic area, time, city, and competition intervention. It’s all moving, and it’s all dynamic. It's filled with trade-offs, riddles, quirks of human behaviour, city dynamic, competitive pricing, hour, day of the week, rain, and natural calamity. It's the most proximate case of seeing a dynamic optimisation situation at a real time unfolding before your eyes.

The work that I did dealt with looking at data very carefully and deciding what’s going on the market and where it might need intervention. It required knowing and doing three things:

• Looking at data, can I translate the data to reality and objectively tell what’s going on in the market?
• Or for the question that I have, do I have the data ready? If no, can I query it out from Uber’s database and present it in a form that answers the question and tells a story?
• What intervention can I make considering the tools I have at my disposal in terms of incentives, communication, and pricing that might make the marketplace more efficient?

These would be brief summary of the work that I mostly did and tried to do. It’s an interesting space to work on -- especially if you love looking at data and seeing the world simulated before your very eyes. It was a lifetime of privilege to watch how the city you live moves akin to the mapped-out veins of the human body.

Uber folks rightly adopted the phrase #UberOn. That’s what I am betting on.