Let’s consider that we are doing this analysis before the launch in a new city, otherwise we should have a lot of data already, so that we can analyse and understand empirically what number we would need and where we can improve.

I think we need to think here about liquidity which is basically having the match between supply and demand.

let’s consider few edge case examples to clarify the meaning:

• imagine we have in 1 district 100 drivers and 1 rider. This would result in excellent experience for the user, but bad experience for the riders, as lot of them would stay without orders
• vice versa scenario could be that there are 100 riders and only 1 driver who could be happy, but 99 users won’t be happy

So I suggest we look at this problem from a marketplace perspective and ensure both actors are getting value out of it.

Define main goals of the launch:

• User is able to get the car within 10 mins
• Driver when finishes the ride already sees the upcoming new order waiting for him. so the driver can be effective and not waste time just sitting on the road.

 

Define the ratio of drivers vs riders. How many riders can the driver serve per shift with steady demand?

1 driver = 2 orders per hour

let’s assume driver works 8 hours: 8 * 2 = 16

let’s assume that every user is unique and the driver serves 16 users per day.

Let’s imagine SF with 7×7 miles area. it results in 49 square miles.

let’s assume every square mile needs a driver to provide quick pickup up to 10 mins.

in this case if we have simultaneous 49 drivers we can serve 49 * 16 = 784 users.

This covers steady demand without peaks. Once the rider base grows, such ratio can help to understand how many more drivers need to be employed more.

For instance if demand grows 2 * 784 = 1568, driver base needs to be grown accordingly: 49 * 2 = 98.

 

Understand the city specifics, user needs and traffic trends

More Factors we need to account for peaks and analyse the trips pattern during high demand time.

 

High demand time:

Working days. Peeks in the morning and eve

• in the morning people want to commute from sleeping district to work/school/university/city centre
• lunch-time people make small commutes for food
• in the eve people want to commute back to home. from city centre to sleeping district
• in the eve people want to commute within city centre to hang out

Weekend. Peeks in the eve:

• Cinema/Shopping malls/parks/walks
• Training activities
• Visiting friends/having parties

Bad Weather:

• much higher demand when people do not want to walk outside

Launch options

Let’s assume we now have 784 users and 49 drivers. Imagine in the morning 100 users want to ride from the sleeping district to work during rush hour. To cover high-demand time there could be 2 Solutions:

1. Keep number of drivers as 49 so we are able to serve steady amount of orders

Trade-off: During peak hours only part of the riders could get a ride for a higher price.

Result: Drivers are happy and constantly employed. Users would not be able to get the ride

2. Increase the number of drivers to cover peaks, analyse hot areas and grow supply.

Trade-off: During peak hours some riders will not be able to get the ride

Result: Riders are happy, though drivers would not be getting enough orders during the steady time.

 

Launch and calibrate:

Both solutions result in either driver or rider impact, so it would require more mechanisms and solution pillars to improve this and actually balance them to achieve better satisfaction for both user groups. Getting more data would help to calibrate the ration and consider some ideas:

• analyse demands per district during peak time to improve the supply
  • employ more drivers in certain hot areas
• during limited supply regulate demand with higher trip price
• trips matching to keep supply in certain areas
• monetisation mechanism to encourage some unpleasant trips, e.g long rides to the district with low demand

 

Approach for the launch:

• set the goal for the launch
• define the ratio between riders and drivers for the initial launch
• understand the city specifics and traffic trends
• launch
• grow and balance the number of riders with available supply. make sure that ratio is not broken and supply and demand are growing proportionally
• calibrate the ratio based on the data