Is Platform Business Model Possible In The Kenyan Banking Market? (Part One)
Part One — The Building Blocks:
It is a well-documented fact that platform-based internet companies have been very successful in the last 20 years. A quick Wikipedia search shows that eight out of the world’s ten largest companies from a market cap perspective are Platform technology companies, only Berkshire Hathaway and Johnson and Johnson are exceptions.
What Is A Platform Business Model?
Moazed and Johnson in their book Morden monopolies describe platforms as;
“… a business model that facilitates the exchange of value between two or more user groups, a consumer and a producer. In order to make these exchanges happen, platforms harness and create large, scalable networks of users and resources that can be accessed on demand. Platforms create communities and markets that allow users to interact and transact.”
Moazed, Alex; Johnson, Nicholas L.. Modern Monopolies (p. 29). St. Martin’s Publishing Group. Kindle Edition.
Moazed and Johnson continue to argue that platforms do two things;
- Reduce transaction costs.
- Enable complementary Innovation.
The transaction cost approach to the theory of the firm was created by Ronald Coase. Ronald Coase describes transaction costs in his article “The Problem of Social Cost” as below;
“In order to carry out a market transaction it is necessary to discover who it is that one wishes to deal with, to inform people that one wishes to deal and on what terms, to conduct negotiations leading up to a bargain, to draw up the contract, to undertake the inspection needed to make sure that the terms of the contract are being observed, and so on. These operations are often extremely costly, sufficiently costly at any rate to prevent many transactions that would be carried out in a world in which the pricing system worked without cost.”
The Problem of Social Cost, Ronald Coase, pg 15 , https://www.law.uchicago.edu/files/file/coase-problem.pdf
To make Ronald’s explanation more clear, transactional costs can be broken down into;
- Search and information costs.
- Bargaining and decision costs.
- Policing and enforcement costs.
It’s therefore crucial in the design of a Platform business to have two things at play:
- Reduction of transaction costs.
- Building in network effects.
1. Reduction of Transaction Costs.
The Uber example.
Uber has taken away or streamlined all the three costs identified by Ronald Coase in implementing a ride-sharing Platform Business Model. The example below shows how one would move from point A to B using traditional Taxi services in a city like Nairobi without an Uber-like service.
If you need to move from point A to B by Taxi, you will encounter the first cost of Search and information.
This involves finding a Taxi that is ready to ferry you to your destination. In a city like Nairobi, you will need to walk to a Taxi bay where Taxi drivers park waiting for customers. Once you are there, you will need to evaluate the type of vehicle that would fit your preference. Here we are talking about the car’s general condition, cleanliness and roadworthiness. You will also make a quick evaluation of the driver. You want to assure your self that the driver looks neat and also make an evaluation as to whether they are trustworthy. Your safety assessment antennae are extremely heightened at this point. You actually depend on what Daniel Kahneman in his book “Thinking, Fast and Slow “ referred to as System one. This process becomes exceptionally complicated at night. You will need to depend on System two at night and only call a trusted driver that has ferried you before. Hopefully, he or she is available.
Once you are comfortable with the vehicle and the driver, the second cost immediately kicks in: Bargaining and decisioning. You will engage the driver in a quick negotiation about the destination and how much it will cost you. If you know your way around, you will have some reference prices in your mind based on distance. This will help you assess whether the price given to you by the diver is fair. If you are new in the city, you will accept any price given to you as you have no way to discover price. To make matters worse, Taxi’s in Nairobi are not metered.
If you don’t reach an agreement with the driver on price, you will have to go back to step one.
Lastly, as you board the Taxi and speed off to your destination, you may not have certainty that the price negotiated in step two will hold. The Taxi driver is also not confident that you will pay them as agreed. Policing and enforcement costs come into play.
The solution to deal with these three costs is achieved by putting in place what is called a Core Transaction. This is the transaction at the heart of a platform. A Core Transaction that is well designed will reduce the three transactional costs identified earlier.
“The core transaction is the platform’s “factory” — the way it manufactures value for its users. It is the process that turns potential connections into transactions. Getting the core transaction right is the most important piece of platform design, as the platform will need its users to repeat this process over and over to generate and exchange value.”
Moazed, Alex; Johnson, Nicholas L.. Modern Monopolies (p. 39). St. Martin’s Publishing Group. Kindle Edition.
To resolve the Search and information costs for the rider and driver, Uber has aggregated verified drivers on its platform to create liquidity network effects. This means you just need to fire up your Uber app select your destination and an Uber taxi will be available to you.
To address Bargaining and decision costs, Uber has implemented marketplace technology that has a real-time algorithmic decision engine that matches supply and demand as detailed in their S1. This real-time algorithmic decision engine technology has three elements;
- Demand prediction: Their proprietary demand prediction engine uses data to predict when and where peak ride volume will occur, allowing them to manage supply and demand in a city efficiently.
- Matching and dispatching: Their proprietary matching and dispatching algorithms generate more than 30 million match-pair predictions per minute. This allows for a rider to be matched with a driver extremely effectively.
- Pricing: Their technology sets product pricing in real-time at a local level. In areas and times of high demand, they deploy dynamic pricing to help restore the balance between Driver supply and Consumer demand. Dynamic pricing helps to balance demand during their busiest times so that a reliable ride is always within reach.
To address Policing and enforcement costs, Uber has designed its platform to optimize for Safety and trust as documented their S1. They have designed their products to include robust safety tools for all platform users. The Safety Toolkit allows both Drivers and Consumers to access a menu of safety features directly from the home screen of the Uber app.
They have implemented a two-way rating system that enables both Drivers and Consumers to rate each other, which increases accountability on the Uber platform.
As we have seen, a platform needs to have a core transaction design that solves the three fundamental transactional costs identified by Ronald Coase. This actually becomes a formula for identifying opportunities in the market place that lend themselves to creating a sustainable platform business.
2. What Are Network Effects?
James Currier & the NFX team at www.nfx.com describes Network effects as;
“…. mechanisms in a product and business where every new user makes the product/service/experience more valuable to every other user.”
The best example of Network effects is one that is depicted by telephone connections. The underlying secret sauce that makes telephone connections have such powerful networks effects is the Metcalfe Law.
A quick search on Wikipedia yields that the Metcalfe law is;
This image from Sketchplanations vividly shows how this law works.
To complement the sketchplanation above, the table below shows how powerful network effects are. The more you add nodes to the network, the more valuable the network becomes. This is due to the possible number of connections possible between the nodes. Once these connections reach critical mass, they have a lock-in effect.
If your business model is designed to create value from these connections, value creation can also grow exponentially.
Unpacking Network Effects — The Safaricom GSM and M-PESA Case Study.
Safaricom PLC is the largest Telecommunication company in East and Central Africa with a market cap $1.364Bn as at 24th Dec 2020. It also has the largest and most successful mobile money implementation in Africa, M-PESA.
I have applied the principles of networks effects to show the underlying forces behind the success of Safaricom PLC as a GSM network and a Mobile Money operator. We will see how these networks have a reinforcing effect on one another from a value creation perspective.
Step One — GSM Network (Layer one and Layer Two):
Safaricom PLC created a successful GSM network that had Kenyan consumers join in droves thanks to Safaricom PLC commercial efforts.
The GSM business is made up of a Physical Network (Direct) of cellphone nodes(Layer 1) and a Physical Network (Direct) of Airtime Distribution Agents (Layer 2).
As we have seen, Robert Metcalfe proposed that the value of the network is proportional to the number of connected users squared ( N ^ 2). As such, the power of network effects gets unleashed once they reach a critical mass activating the lock-in effect.
In Safaricom’s case, this lock-in effect set in as the network grew, making Safaricom, the largest Telco operator in the country. Kenyans were therefore disincentivized from joining competing networks because everybody they knew was on the Safaricom GSM network. Calling them from any other network was expensive. Switching costs became too high for customers.
Step Two — M-PESA Mobile Money Network (Layer Three and Layer Four):
Safaricom PLC layered on a Mobile Money Network. This is a Personal Utility (Direct) Network (Layer 3) that sits on top of the Physical GSM Network discussed in step one. As part of implementing the Mobile Money Network, they also added a Physical Mobile Money Agent (Direct)Network (Layer 4).
The Physical Mobile Money Agent (Direct)Network provided cash liquidity for the M-PESA Personal Utility (Direct) network nodes, allowing for cash in and cash-out transactions to happen therefore creating Utility for the Mobile Money Network at large. The M-PESA Mobile Money network grew in scale very quickly.
The diagram below shows exactly why this is the case—four networks working in concert with very powerful network effects at play.
Step Three — Lipa Na M-PESA Network (Layer Five):
Safaricom went ahead and layered on a fifth network. If you think about it, this is quite impressive.
Safaricom added a Two-Sided Market Network (Indirect), the Lipa an M-PESA Network (Layer 5) that facilitates payments between buyers (Consumers on the Mobile Money Network) and Sellers (On the Lipa Na M-PESA Network).
To activate the indirect links between sellers and buyers on the Lipa Na M-PESA network, Safaricom made merchant acquiring costs cheaper than the competing networks that were promoted by Visa and MasterCard. Due to the prevalent multi tenanting on the Merchant (seller) side, Merchants had no problem accepting M-PESA due to the cost-benefit the Lipa Na M-Pesa network offered. Again this network grew very quickly to become the dominate Merchant acceptance network in the country.
With network effects, the value is in the network and not in the product. As an example, the M-PESA product features have been unchanged for more than 5 years but they still continue to offer massive utility to customers. See below;
- Moazed, Alex; Johnson, Nicholas L.. Modern Monopolies (p. 29). St. Martin’s Publishing Group. Kindle Edition.
- Kahneman, Daniel. Thinking, Fast and Slow. Penguin Books Ltd. Kindle Edition.
These are my humble observations, and I leave the door open to the fact that I could be totally wrong about different aspects of the Platform business models. Let me know what you think about Platform business models and what your unique insights are.
The opinions presented here are strictly my own and do not represent those of my employer.