In 2011, Eric Ries popularised the concept of a Minimum Viable Product (MVP) as part of the Lean Startup methodology. An MVP can be defined as a product with just enough features to attract early adopters and gather feedback for improvement. According to Ries, building an MVP helps startups only invest heavily in a fully developed product after first confirming whether it addresses a real market need.
To what extent can this approach be transposed to digital platforms? To address this question, we follow the analysis presented in our book and proceed in two steps. This article shows that we cannot simply substitute ‘platform’ for ‘product’ in the MVP acronym. In the following article, we will question how minimum a platform should be to be viable and whether the minimum viable platform shouldn’t be a pipeline to start with.
From “Product” to “Platform”
At first glance, extending the concept of MVP might seem appealing by interpreting the ‘P’ as ‘Platform’ instead of ‘Product’, as Choudary (2014) suggests. A ‘Minimum Viable Platform’ (MVP) would then be defined as the simplest version of a digital platform, equipped with just enough features to facilitate the platform’s core interactions, allowing it to serve its primary purpose while being able to gather feedback from users for further development.
This analogous concept certainly offers significant advantages. For instance, Rachitsky (2019) provides empirical evidence that many successful platforms began their operations with an MVP. That is, they initially limited their operations by focusing on a specific category of products or services and/or a specific geographic area. It is only at a later stage that they decided to expand their operations, guided by feedback from early users. The table below illustrates this point with three well-known examples.
From his analysis, Ratchisky concluded that “The best way to get big is by first going small.” He did not claim, however, that going small is the optimal way to succeed. There is indeed one discordant voice in the interviews that he conducted: Sander Daniels, a co-founder of Thumbtack, told him that while conventional wisdom suggested narrowing the focus to a specific category or geography, Thumbtack did the opposite. They covered all categories and geographies from the beginning. This broad approach increased the frequency of use and allowed them to scale rapidly, providing the revenue and user base needed to build a successful platform. In their case, a narrow focus would not have worked.
One could say that Thumbtack is the exception that proves the rule. However, other cases confirm that focusing on a restricted set of well-identified features may not guarantee success. For instance, Quibi failed quickly after its launch in 2020. This short-form streaming service struggled to gain a user base and retain subscribers due to a combination of factors, including a lack of compelling content, an overestimation of consumer demand for mobile-only streaming content, and an ill-timed launch during the COVID-19 pandemic.
A more recent example is Artifact, the news app from Instagram’s co-founders, which announced its shutdown in January 2024, less than one year after its launch. Despite its well-received design and AI-powered features, it struggled to expand its user base outside the U.S. and saw a steep download decline after its initial launch. In a company blog post, co-founder Kevin Systrom explained that the market opportunity was too small to justify further investment.
Thus, caution must be exercised when applying the MVP concept to platforms. The main reason is that, unlike products, platforms do not offer intrinsic value; instead, they co-create value with their users. This affects the concept of MVP in two significant ways:
- The V conditions the M because a platform must attract a minimum number of users to be viable;
- The V also conditions the P insofar as the only path to viability may be to start as a pipeline and only turn into a platform later.
We examine these two issues in the following article in this series.
(During the preparation of this post, the author used genAI tools to collect ideas and improve the expression. After using this service, the author reviewed and edited the content as needed and takes full responsibility for the content of the publication.)
