Most important steps in platform businesses

Businesses like AirBnB, Uber, eBay, Alibaba, Tinder boom in recent years. More and more platform businesses rise from the ground.

Though the efficiency of these business models are proven if they scale, but is it truly the appropriate model for any industries? What are the most important factors to keep in mind when building a platform business?

In this article I give you solutions for these questions. I would like to emphasize this is an introduction to platform business models, a quick summary to help you understand the basics of this business model.

1. What is a platform business model?

We used to live in a business environment where commerce streamed linearly (pipe businesses). Companies added value to their products, managed the logistics and sold them to consumers. Producers and consumers held completely distinct roles. By the appearance of multisided platforms it has changed.

Multisided platforms are technologies, products or services creating value mainly by enabling and enhancing direct interactions between two or more customers.

Platforms have distinct advantages over pipe business models, such as more favorable cost structures, higher profit margins and the ability to scale to a size that pipe businesses can’t. Platform businesses capitalize the network effects and reach vast growth if they can scale it.

In case of all platform businesses there are at least 3 participants: supply side, demand side and the platform provider. The platform providers make sure that customers can attract, match, connect and transact within the platform smoothly. The suppliers are the producers, sellers, hosts of the specific product while on the demand side are the consumers, buyers, guests of the product.

2. Types and examples of platform business models

Without covering full scope of platforms, there are:

  • Technology platforms (Amazon Web Services, Microsoft Azure, Twillio),
  • Computing platforms (Android, iOS, Windows),
  • Utility platforms (Google Adwords, Google Maps, App Annie),
  • Interaction networks (Facebook, Twitter, LinkedIn),
  • Marketplaces (AirBnB, Kickstarter,,
  • On-demand service platforms (Uber, Amazon Home Services, Munchery),
  • Content platforms (YouTube, Medium, Pinterest).

These businesses are different in many aspects but all of them have similar characteristics. They all have platform provider, supply and demand side. They all face the chicken and egg problem (which side to recruit first), they all have to find the perfect ratio of the supply and demand side (though the ratio itself differs in all cases) and need to find the appropriate timing, they all benefit from the network effects and so on.

3. Is really the platform model you should use?

If platform businesses scale and utilize the power of communities they can eventually print money.

It is easier said than done, so first of all let’s see if you should really use the platform model for your business. To decide first you need to answer the following questions:

  • Is there any demand for your service? Are they engaged enough to pay? Well, it is relevant for all kind of businesses so start with this question.
  • Are there two sides of participants (supply and demand side)? It is important to examine if you have both producers and buyers.
  • Is there actually supply? Are the producers motivated enough to produce? To answer this question it is important to look after what they currently do, are they easily convincible to produce on your platform?
  • Will the ratio of supply:demand be ever scalable? In other words is the amount of your potential suppliers enough for the demand side and vice versa?
  • Is there repetitive demand on your service or is it a one time demand? For example in case of Uber guests use it frequently, but in case of high price products consumers mostly buy it once in a while (e.g.: real estate, car etc.)
  • Is it easy for your customers to avoid your platform and transact outside of it? Can you apply any kind of barriers that makes it difficult for both parties to avoid you? Are they even motivated to avoid you? If you use transaction fees as your business model they are most probably motivated to do so.

4. Most important steps to keep in mind when building a platform business

If you got positive answers for your questions in the previous chapter then you made your first validation about the viability of your business idea. Now it is time to work on your so called rocket model. Almost all platforms need to:

  • Attract a critical mass of each segments of customers (supply and demand side),
  • Match them,
  • Make them connect,
  • Let them transact,
  • And eventually optimize the whole process.

What you need to keep in mind is that these steps vary based on the maturity of your platform: you have different attraction strategy when your business is in the design phase, ignition phase (finding product-market fit), scale-up phase, or the maturity phase. I will only cover the most important factors in general. If you are interested to learn more I recommend reading Platform Strategy by Laure Claire Reillier and Benoit Reillier.

For a vivid platform you need to attract both your supply and demand side and need to reach a critical mass. In this phase you need to consider which side to recruit first? Often, it is the supply side you need to pursue first. You need to figure out how to retain them until buyers come. E.g.: Uber payed their drivers a basic fee just to be available until passengers started to use it in the city they launched.

What is the amount of suppliers when you can start attracting the demand side? What is the supply:demand ratio that your platform can work with? In other words, how many suppliers are needed to attract buyers, and how many buyers are needed to employ suppliers.

After attracting both sides of customers you should introduce and match them. You as a platform provider should enhance suppliers to introduce them on the platform, make it easy for both sides to provide the necessary information about themselves.

Furthermore, the matching function is key to maximize positive network effects. It has to meet participants’ needs, be timely and provide the right amount and depth of information. It needs to provide the optimal amount of choices leading to successful connection. Matching function can be done in several ways: with search functions (like in most product marketplaces: eBay, Amazon Marketplace etc.), geo-location (Uber, AirBnB), machine learning based algorithms based on your behaviour on the platform (Tinder, Facebook etc.).

Mostly, people would like to know more about matching with their results. In case of a low value standard products (like buying a T-shirt on eBay, or ordering an Uber) connection phase is skipped. But in other cases connection is helpful to know more about your result before transacting.

The biggest challenge in this phase is to provide the opportunity to connect with each others so they have the right information to decide about transaction owning the risk that you will be avoided as a platform and transaction will be made outside of you.

You have many tactics to apply in order not to be jumped over: block sharing personal information (AirBnB blocks it in their messaging function), provide trust, easy payment, refund policy, customer support. You can also build a model when you are unavoidable (like Uber), but in these cases no connection is needed.

Customers are motivated to avoid you if you use transaction fees as your business model, think about applying other methods to generate revenue: freemium model (make it free but charge for extra services), subscription model (LinkedIn), or do not care too much about charging your users but use their data, analyze it and sell behaviour data for giants.

Transaction does not always come with revenue, it differs from platforms to platforms: clicking on a Facebook Ad, ordering an Uber, going on a date, downloading an app from App Store, hiring someone on LinkedIn are core transactions. In this phase our goal is to maximize the value of our core transactions even if it is instant revenue or a transaction (mostly behaviour data) leading to revenue later on or rises the value of the platform.

Finally, the most important part is to optimize your whole process. Collecting lots of data on each stages help you analyze the bottlenecks and the management can work on to fix them so you can truly unlock the potential of network effects. These bottlenecks can be:

  • Not finding the supply:demand ratio in the attraction stage,
  • Wrong search algorithm in the matching stage,
  • Too many leakages (customers avoid your platform) in the connection stage,
  • Not enough core transactions in the transaction stage.

I only listed one example per stage, but in reality there are always many bottlenecks you need to fix. It is a constant iteration and improvement based on your actual performance.

In conclusion, it takes time and consideration to decide whether you should build a platform business model for your idea or just go with a pipe model. It is always worth to go through the rocket model in each phase of the project (not only in the design phase, but in the ignition, scale-up and maturity phase as well), because it can save you lots of time and money.

Whether you consider to build a platform business, or already did one, or even just interested in the topic, feel free to drop me a line on LinkedIn. I would be more than glad to discuss any concerns.

If you enjoyed this article please recommend and share.