3 Key Metrics to Manage Complex Marketplace Businesses. Lessons from Uber and Toptal
November 29, 2021
Uber is one of the strongest operational companies out there that has cracked how to scale marketplace business. Their secret was simple - a launch playbook and use 'health' metrics to monitor supply and demand match.
Let me share what those are and how to apply this your marketplace business.
Prior to my stint at Toptal (a unicorn talent marketplace) I was working for a PropTech marketplace Property Partner (acquired by Better). We had quarterly workshops with Niall Wass COO of Uber Europe (at a time) who shared how Uber has built a system to launch 2 new cities every month.
Uber's Playbook To Launching New Cities
Aleksandr, CEO of Lemon.io has quizzed 5 early employees about Uber's launch playbook. I've borrowed his summary below:
- There is a city launcher that moves from city to city every few months [Alex: you can have many of these operating at a time]. Their task is to hire people and launch operations.
- In most cities, they hire a general manager, a marketing manager to drive demand, and an operations manager for supply acquisition.
- Every team acts independently from the core team. They optimise for speed, use their own resources and make their own decisions.
- Every city has its own metrics, and there is no excuse not to deliver.
- Even though they were scaling fast, operation managers spent enormous time with drivers, held events, and onboarded them manually (sometimes in person). It gave them a ton of insights, helped to build trust, and bring even more drivers through referrals.
- Uber has a well-written manual on how to launch cities. It is so good, even a newbie can do it.
- Yes, there were mistakes and problems, but if optimise for speed and grow so fast, those mistakes become irrelevant.
The above process describes well how to launch new operations, but maintaining and monitoring them is a whole different ball-game. Especially, when there is both supply and demand to worry about.
Niall has rolled out what he calls 'Marketplace Health Metrics' that Uber has used to successfully scale their operations after launching.
What Are Marketplace Health Metrics?
A set of KPIs to measure health & satisfaction of both supply and demand sides of a marketplace. These metrics aim to:
- Track the main drivers of growth on both sides of the market
- Provide focus to internal stakeholders
- Easily track and understand where each specific market currently sits in terms of performance and lifecycle (assuming operations in multiple markets i.e. territories)
These were 3 core metrics on each side of the market: demand and supply. The numbers below are just examples to illustrate the point, they are not real Uber numbers.
On the demand side (riders):
- More than 90% of users who got a ride within 5 minutes (main job to be done for user and it means there are enough drivers, otherwise user satisfaction will go down)
- More than 80% of users who successfully completed a ride on their first try (no IT glitches or drivers cancelling rides - otherwise bad user experience)
- Users, on average, paid 20% less than an equivalent trip in a taxi would cost (financial incentive to use Uber)
On the supply side (drivers):
- Drivers, on average, have utilisation of 80%+ (earning a living is a main job to be done for the drivers)
- Driver NPS > X (satisfaction is very important to retention)
- Drivers, on average, made 20% more than average local salary index (financial incentive to use Uber)
If the metrics are 'healthy', above set threshold, the marketplace is balanced and will grow.
You can see the pattern:
- Metric 1 - ability to complete their main job to be done (JTBD)
- Metric 2 - user experience check
- Metric 3 - financial incentive to use the platform (and grow it)
In Uber's case, their main growth channel was virality and word of mouth. Ensuring good experience is key to maintaining it.
However, if the metrics are not balanced you need to lean into it and fix that part. For example:
- If only 60% of users can find a driver within 3 minutes you need to step up your driver acquisition.
- If driver utilisation is 50% you need to find more riders to bring it back in balance.
- If driver NPS is low, you need to improve their experience before they churn
Being brutally focused is the only way to manage a complex business like a two-sided marketplace. Otherwise, you'll be fighting windmills all day, every day ;-)
Applying Health Metrics For Your Marketplace Business
Uber has used this with a great success in their business but this can be applied to any business, marketplace or not.
Here's how I would establish this framework for a talent marketplace business like Toptal or Andela.
In their case, there is no concept of cities or countries as they have globally distributed talent that works remotely for their clients. The equivalent in this case is the specialism of their talent, for example front-end developer or iOS developer. That's where they need to balance their supply and demand.
Here are the metrics I would use in a talent marketplace (the numbers are just examples):
Demand side (clients-buyers):
- More than 90% of clients start their projects within 3 days (their main job to be done)
- More than 90% of clients ranked their freelancer 4/5 or higher in their latest survey (could be done together with an invoice) (quality and satisfaction)
- More than 90% of clients got their freelancer for less than the local market rate (financial incentive)
Supply side (freelancers):
- More than 90% of freelancers got their job within 30 days from declaring their availability for their requested hourly rate (their main job to be done)
- Freelancer NPS (user experience)
- Freelancers, on average, have utilisation of 75% or more of their allocated time (financial incentive)
If you monitor these metrics by specialism, you would know:
- If you need to find more iOS developers to prevent supply crunch.
- If they are not being hired fast enough you need to step up your client acquisition efforts.
- If both side are satisfied, they will continuing using your platform increasing their LTV and your revenue.
This method can be applied to any business and enables you to monitor complex, multifaceted operations. First, you need to establish at what level do you face supply and demand issues. In Uber's case it was cities, in Toptal's case it was specialisations.
Then you need to identify key metrics and thresholds that would ensure a smooth and healthy operation.
If you get below threshold you would need to address the issue and bring it back in balance.
It's simple, but not easy.