Platform Games: Part Deux

Lockdown is working great for writing! No distractions.

Let me jump into it straight away with an entrepreneur's words of wisdom as he wound down his company:

“The initial seven years were all about having negative working capital, positive cash flow, and a sustained ability to fund our own growth. Those were the only metrics we tracked. In the last 3–4 years, though, I can honestly state that, somewhere, I lost my path. I started treasuring GMV, room-nights and other ‘vanity’ metrics instead of the fundamentals of cash flow and working capital.”

The digital platforms we discuss here are transactional platforms (Amazon, Uber, Netflix, Urbanclap, etc). They aggregate supply and demand in imperfect markets that are craving for new forms of engagement. 

How does a digital platform differ from a ‘value-chain’ business? I recommend Stratechery by Ben Thompson for a deep discussion. Value-chains thrive on backward and forward integration. They integrate suppliers and distributors via exclusive relationships to gain command in the market and create competitive entry barriers. Customers are important but come last amidst all the other activities. 

Digital platforms disrupt this model. A platform provides a space where various stakeholders come together and create value to each other. There are four specific differences between platforms and value-chains: 

Modularization/Disaggregation of Suppliers: Platforms disaggregate the supply-chain. The components of the value proposition are independent (and numerous) but function as a whole from the consumers’ perspective. E.g., Netflix aggregates disparate pieces of content that would not be at the same place under regular circumstances. More importantly, the content is available to the customers at any time and in any order. 

Network effects and Digital Scale: What matters is the experience of the end-consumers. As more consumers have great experiences, the more they use the platform, hence attracting even more consumers and suppliers. It creates virtuous network effects that solidify the gains. Since marginal costs are minimal, scaling a platform is theoretically a less-capital intensive effort. The term that entrepreneurs love to describe these gains is ‘winner-takes-all’. In the process, the suppliers become commoditized, and have to compete with many others for attention and these many others are geography agnostic. The best part: whatever happens on the platform, the platform developer gets a cut. 

Digital Standardization and Disintermediation: Platforms control the interactions via digital standardization methods such as data-based algorithms, reputation systems, entry protocols and payment mechanisms. This eliminates inefficient and dispersed intermediaries/suppliers. E.g., TripAdvisor created a UGC driven reputation system to digitize a significant part of hotel booking process. Therefore, it did not matter even if the hotels did not want to be on the platform. 

Data-driven Personalization: Since platforms own the consumer-supplier interactions, they own vast amounts of data that can be used to better understand consumer needs. They can personalize products and user-experiences, and hence create efficiency in the supply-chain. This creates a customer lock-in that eventually becomes the entry barrier for others in the market. E.g., Facebook enables user interactions, and captures substantial data about their behaviors. Its big promise to advertisers is micro-targeting at scale. Advertisers love to hate FB but they cannot live without it. 

Fabulous right? with several caveats! To achieve these outcomes, platform builders have to deal with 4 critical problems:

Problem 1: The value proposition and market size

Platforms fail when they play in markets that are not ready i.e., the need is not strong or large enough. Platforms typically thrive by eliminating an existing imperfection/friction that vexes consumers and suppliers. The attractiveness of the platform is a function of { a) extent of the friction, b) how the digital solution eliminates entire or part of the friction, and c) someone is willing to pay for it} Here are two examples to illustrate the challenges. 

Solution: There is no single solution except for intelligent market research, and test and evolve methods. Nothing better than good old leg work in the market. 

Problem 2: Seeding model (Chicken or egg problem)

This can make or break a platform. The problem is simple: without suppliers, buyers won’t care about a platform. A platform without sufficient number of buyers if is even less attractive to the suppliers. 

Many platforms use brute force (splurge money/provide incentives) to seed the platform. It could work but is a bad idea in the long-term since customer acquisition costs go out of control, and eventually kill the platform. Certain amount of incentives are necessary. But it is important to strengthen the value proposition in the seeding stage. The ultimate outcome of this stage is to create a critical mass of initial users that transact with each other. Eventually the network effects kick-in. 

Solution: I have seen 4 smart ways to solve the problem.

Problem 3: Governance Model

Governance model determines the composition of, and the extent of value creation  on a platform. Many times, platform developers take these decisions in an ad-hoc manner. Their strategy is to increase the numbers as much as possible without realizing that it can debilitate the platform. Three questions to ask:

Who should participate? - Define standards and requirements of participants

What should they do? - Define the activities they need to perform to create value

How will they operate? - Define rules of engagement including investments and extent of relationships

Based on these key questions, platforms have a choice between an open or a closed model (or something in between) 

Eventually, we see that many platforms move from a closed to an open system as they scale.  Good example is Facebook – It was initially restricted to Harvard students but subsequently opened up to all others. 

Solution: How does one pick a model? It helps to consider two inter-linked dimensions:

a)     Is the strategy about quality or quantity? If quality, closed system e.g., Uber. If quantity, then open e.g., YouTube

b)    Is the value proposition producer driven or network driven? If Producer-driven, closed e.g., iTunes. If network driven, then open e.g., FaceBook

Problem 4: Revenue model

Studies show that a wrong revenue model is one of the top reasons that platforms die. Even platforms that burn tons of VC money need to consider making money at some point. The economics have to work out eventually. 

It is also one of the most challenging decisions since the platform developer has to manage three competing motivations: 

A)    Grow the network on the platform: If the fees is too high, the users have a motivation to skirt the platform and form direct connections. 

B)    Enable all the participants to derive value (could be non-monetary as well)

C)    Extract economic value for itself (monetary!)

D)    More importantly, the revenue model has to be structured in such a way that it strengthens the network effects. That means the side of the platform that brings maximum network effects should ideally be charged less. 

Solution: Each of the above is a non-trivial outcome. There is no single formula to define the revenue model of a platform, and platforms are known to use a combination of models depending on their strength. 

 In my opinion, two dimensions are of utmost importance while discussing the revenue model:

-       Whom to charge? (Customers or suppliers)

-       What to charge for? (access or transactions)

Here is a 2X2 matrix that can provide some thumb rules. It is not exhaustive. 

I think I have written enough! 

In article 3, I intend to highlight 5 pitfalls/myths/legends that are pervasive in the startup world and how to think about them while building a platform strategy. Any comments or suggestions are welcome.