This was by far my favorite talk.
Two sided markets are everywhere. Two examples are bank cards and merchants, and content producers and content consumers. However, there are not any good frameworks for evaluating the optimal growth strategy for markets that exhibit this behavior.
The presenters set out to create an applicable model. The first step was analyzing the types of variables that would go into this model. Since they were approaching the problem of expanding a payments network, they chose to focus on merchants and card holders.
They settled on a few key metrics, such as cost of acquisition for a card holder, the cost of acquisition for a merchant, and the number of each side that would have to exposed to the service before converting to use the service. They modeled the transactional behavior of card holders and merchants with two power law graphs.
They decided to explore a few different strategies using this model over the course of a set number of rounds. The first three were one sided, focusing entirely on either consumers, low-volume merchants, or high-volume merchants. They also tried a few mixed strategies, alternating between consumers and low-volume merchants, or consumers and high-volume merchants.
With their example input, focusing on consumers was the correct course of action. However since the inputs are variable, one could run extra simulations to find the breaking point, or adjust it for real world feedback, such as a different acquisition cost.
Since so much of building a market is guestimation, it would be wonderful to have access to a configurable model such as this. If they make the code available, I will definitely be adding some R skill and playing with it.
Did you like this? Please share:
The Lost Year: A Failed Experiment to Switch Away From Mac
Fed up with the Apple Keyboard, I bought a ThinkPad, installed Linux, and promptly decided that I hated computers.
Maker's Space, Manager's Space
The Grand Remote Work Experiment: A Retrospective
The COVID-19 pandemic has lead to an unexpected experiment in remote working. What has worked and why?