The pillars of a platform business

When it comes to platforms, the concept is often nebulous. Many understand it as technology or infrastructure. But the objective here is to talk about Platforms as a business model. In this sense, Platforms has as its central concept to be a model where value is generated by facilitating interactions between one or more parties. Overall, they create environments where these exchanges have lower coordination costs, more convenience and more security.
The most evident model of a platform are marketplaces, connecting producers and consumers of products and services. However, under this “umbrella” there are other platform models:

  • Interaction platforms such as social networks, connecting “producers” and “consumers” of posts, tweets or content in general, for example.
  • Data platforms, where the interaction between each user takes place through the generation and consumption of data that the platform coordinates. Like Waze for example.
  • And the development and infrastructure platforms that facilitate the creation of new applications, usually “on top” of existing products, such as AWS, Salesforce, IOs, Android, among others, facilitating interaction between developers and users. They often modularize the infrastructure so that businesses can scale more easily.

Each of these models has its peculiarities. However, they all have some basic pillars in common, which we will see below. Every startup or company that is building a platform, as a business model, should focus on them from the beginning of their modeling.

1-) Strategic design
Every platform depends on its interactions being scalable to get traction. And this is the function of strategic design: Defining and designing the interactions that must take place between the parts of the ecosystem, with as little friction as possible. And by interaction, we are talking about a relationship between 2 entities, with a unit of value being exchanged (product, money, reputation, quality etc.), the channel and context in which it takes place.
The strategic design of a platform must be composed of 2 major “engines”:

1-) Interaction engine: These are the interactions between users (“peer-to-peer”). Their main objective is to reduce the cost of coordinating search and selection, promote trust and generate feedback. This engine must ensure that the exchange of the Platform’s value proposition actually takes place among its users.
2-) Learning engine: These are the interactions between the user and the platform (“peer-to-platform”). These interactions, on the other hand, aim to ensure support for the execution of “engagement milestones” and the evolution of users within the platform. Every platform must be concerned with how its users evolve within it. Both consumers, but mainly producers. It’s hosts becoming superhosts on Airbnb, it’s Amazon and Mercado Livre’s loyalty programs and delivery apps, for example.

In addition to the importance for its scalability, interactions will also generate a large part of the data that a platform will use for its governance and for generating a network effect. Therefore, they must be designed considering this role as well.

Strategic platform design is about interactions. Your ability to scale depends on them. That’s why it is considered the pavement of its construction.

2-) Governance
Fairly dividing the value generated by a platform among ecosystem participants and ensuring that it will generate positive interactions for everyone involved are the main objectives of Governance.
Especially the producers of a platform will only increase their engagement if they realize that the relationship between their effort and the part of the value generated by the platform they receive is fair. A slight imbalance in this relationship and conflicts appear. The classic situation of this conflict happens when the producers have the platform as their main sales channel and at the same time the platform itself acts within it as another producer. The recent clashes between Apple and Spotify and some of the lawsuits that Google and Amazon have been suffering in court are the best known examples of this.
In addition, we have seen big brands, such as Nike and Ikea, leave large marketplaces like Amazon because they believe that what they add value to the platform is not properly retributed.
Another fundamental role of governance is to regulate that the interactions that take place within it are always as positive as possible for everyone involved. Quality, fraud, security are all aspects that a platform’s governance must take care of. Generally, the curation mechanisms most used by platforms are:
1-) Manual curation: When the platform PM himself curates the interactions. We see this happening for example on Trip advisor when there is an offensive review.
2-) Community curation: When all sides of a platform self-regulate, ensuring the quality of delivery and consumption of the value unit. It is the mechanism that for example Airbnb uses to allow hosts and guests to evaluate each other. It is also the most “scalable” mechanism in marketplaces, mainly.
3-) Automated curation: When certain parameters are stipulated, an algorithm defines the interactions that should be banned. This is the mechanism most used by interaction platforms such as social networks.

Governance is an ongoing process of fine tuning where the main challenge is the optimal balance between maximizing positive interactions and minimizing friction for its execution by users.

3-) Network Effect
Network effect is the movement that makes a product or service become more valuable as more users adopt it. This phenomenon is nothing new. This is what happened with the telephone, credit card and other products and services of this nature. For a platform, it is capable of making each new user generate more value for those who are already on it. The ability to generate network effect is the main growth engine of a platform. It is estimated that 70% of the value generated by platforms comes from it.
In order for the network effect to happen, positive loopings of value generation must be designed for everyone involved in the platform

The more developed the ecosystem and its interactions, the greater is its network effect generation potential.

The network effect on a platform happens from the moment it has “solvency” and enough critical mass for producers and consumers to complete their interactions, which is called the liquidity point. From that moment on, each new user who enters the platform generates more value for everyone, exponentially.

This movement is what allows some marketplaces to completely dominate some markets, or part of them.

Platform network effect brings 2 big benefits in addition to growth. A high scalability power and a great defense of its model against competitors and substitutes.

4-) The commercial launch (The “chicken or egg” dilemma)
The initial traction of any business is always challenging. But platforms have a particularity in this aspect. Producers will only engage in the platform if there is demand, and consumers, in turn, only if there is supply. How then to bring both sides simultaneously and generate the necessary traction? This is called the “chicken or egg dilemma”.

Platforms that overcome this challenge generally followed these 5 steps:
1-) Restrict the market by geography or category: The focus accelerates reaching the liquidity point for the initial generation of the network effect. Facebook in Harvard, Airbnb in San Francisco, Amazon with books…
2-) Prioritize one side of the platform: Identify which side, producer or consumer, has greater importance relative to the other or is more “scarce”. Eventually, it may be the “subsidized” side to enter the platform first.
3-) Leverage the prioritized side: Develop channels and “hacks” to leverage the prioritized side. Rappi in São Paulo, after selecting a neighborhood, manually uploaded its menus to the app. After having a minimum number of orders placed, they visited the merchants, showed the numbers and proposed the formalization of a partnership. Another strategy used to leverage one side is to develop a SaaS for one side that generates engagement regardless of whether the other side exists on the platform. “Come for the product, stay for the network”.
4-) Leverage the second side: With the prioritized side inside the platform, develop the channels and actions for the non-prioritized side.
5-) Leverage the platform and its functionalities for both sides, break the liquidity point and trigger the generation of network effect.

Until reaching the Liquidity point, the platform must have a clear strategy to obtain a critical mass of producers and consumers, simultaneously. Being able to estimate the time to reach the liquidity point is essential to know what will be your “burn rate” and the financial health of your business. This is the biggest challenge at the beginning of any platform.

5-) Monetization
None of the above pillars will make sense if the platform fails to monetize. To define your monetization model, you need to answer two questions: First, charge from whom? And second, for what?

Now, to answer each of these questions, it is necessary to understand the particularities of each platform submodel that we talked about at the beginning of the text.
For example, for marketplaces, we generally assess the level of complexity of interactions. If the value chain being “platformed” is complex, interactions naturally arise that can be monetized as “premium” functionality. Another important aspect to evaluate in marketplaces is the value that one side has in relation to the other. If one side is relatively more valuable to the other, it’s possible that a “paywall” model will fit in to access it. Now if we have a model where interactions are not complex and the relative value between the parties is very symmetrical, we inevitably fall into the commission model, the so-called “take rate”, as in the case of Uber.
As for interaction platforms, we usually analyze the nature of interactions, whether it is “1 to 1”, or “1 to n”. And the network density, that is, the total potential of network interactions, how many can actually be connected on the platform. Generally social networks have a “1 to n” interaction nature and a high network density (you can theoretically connect with everyone), which favors models with advertisers. On the other hand, platforms with “1 to 1” interactions and low network density, the “fee” per transaction ends up being the most adherent model, such as Paypal, for example.
Obviously these model possibilities can coexist as the platform and its model evolve.

The definition of the most adherent monetization model depends on an evaluation of the characteristics of the platform’s network and its business dynamics.

The pillars of a platform business actually behave much more like engines, as one directly affects the other. Both positively and negatively.

This construction is a journey begun in strategic design, but without ever losing sight of the pillars that are to come and the impacts they will have. These pillars, obviously, are not a guarantee of success, as market, execution, team and agility continue to be fundamental in the game. But the platforms that were successful did their homework on each of them very well.

Fernando Teixeira da Silva

More than 15 years in commercial development projects. Passionate about commercial strategy, business models and go to market strategies.