The evolution of the Open Source market & how Fair Code is taking over
The world of Open Source software has undergone a radical transformation since its inception. Once considered the cheaper, lower-quality alternative to closed-source businesses, it’s becoming the best way to build software in numerous verticals.
Open source is now everywhere with 92% of enterprise applications containing Open Source libraries.
But while 10 years ago, most VCs and observers of the market were quite skeptical about the way to monetize Open Source, new business models, and less permissive licenses (read more about licenses that are a pillar of the Open Source world in this article) allowed these companies to build great businesses.
It caused a sharp increase in VC funding (as you can see in the graph below), and the emergence of market-leading companies like Elastic, Hashicorp, MongoDB, Databricks, etc.
Here, I’ll try to trace the contours of the evolution of Open Source, navigating through its phases, and try to project where it’s going.
The Genesis of Open-Source
The dawn of open-source software marked a significant deviation from established norms. In the beginning (1980s), Open Source was a revolt, driven by developers who challenged the monopolistic dominance of tech giants like Oracle or IBM. They believed that software should be democratized — free, accessible, and modifiable.
Did you know that the fundamental protocols and technologies underlying the internet could have been proprietary and controlled by a single entity? The Internet has its roots in the ARPANET project developed by DARPA. And when Vint Cerf and Bob Kahn developed the TCP & IP protocol they could have made it a proprietary one!
Imagine having one corporation in control of the Internet… 😨 But would the Internet be based on the TCP & IP protocol if it were closed source? Nothing is more certain!
This open era fostered global collaboration among developers, producing great projects like MySQL and Linux. Most of these projects were completely open with licenses such as the MIT license allowing developers to use the code for most use cases, including commercialization.
But while the spirit was predominantly non-commercial, some companies like Red Hat found a viable business model by offering support and professional services. At this point, the ‘service’ model was the primary (and often, the only) revenue stream for Open Source companies.
Fast forward to 2018 when Red Hat is sold to IBM for $34 billion. As of today, it remains the biggest exit for an Open Source company and one of the few companies that succeeded in scaling this business model. Indeed, with low margins, almost no barriers to entry, and the need for a human consultant each time you acquire new clients, this business model is far from being VC-compatible
It’s why, a lot of Open Source companies failed to generate revenues in the early 2000, and some VCs were reluctant to even hear about Open Source.
Open Source 2.0: From Service to SaaS and Open-Core
While in the 90s, open-source developers and big tech companies were engaged in open war, Open Source 2.0 symbolized maturation.
More and more Big tech companies became advocates of Open-Source, releasing some great and successful projects like Kubernetes and TensorFlow (released by Google) without any primary intention for monetization.
On the other side, new businesses emerged adding proprietary elements on top of Open Source projects marking the birth of a hybrid model, from ‘service’ only to the SaaS and Open Core models.
The SaaS business model
It’s a hosting business model. Companies using it have an Open-Source product that anyone can download, modify, and self-host, but they also offer a native hosting solution on their own server for a monthly fee. It makes it easier to maintain the product for technical teams. And for less technical users, it simply allows them to use it (they couldn’t really do it on their own before).
Advantages: Good margins & quite simple to operate
Issues: You can quickly get competition from the biggest cloud vendors that will integrate your service into their offering if it’s successful (e.g.: AWS integrated Elastic services in their offering and stole a huge chunk of potential & existing clients)
The Open Core business model
This business model which is quite particular is gaining traction amongst Open Source founders. With Open Core, companies keep the core of their product Open Source (often > 80% of the code base) but charge for extra features.
The paying features are targeted to the biggest customers. The goal is to keep the product free for individual developers in order to foster community growth while making pay big companies to create a healthy business.
Advantages: Best-in-class margins (>80%) & often bigger share of Entreprise accounts
Issues: This business model is often complicated to operate because it relies on a fine balance between working on the paying features without letting the Open Source part die.
From this phase of Open-Source 2.0 emerged some of the best Open Source companies out there like Elastic Search, Databricks, Cloudera, and Hashicorp.
In the beginning, most of these companies only relied on the SaaS business model.
But while they grew, they experienced one of the biggest challenges of Open Source: the competition from Cloud vendors that basically provided the same services. These cloud vendors like AWS, offered integrated hosting for products developed by Open Source companies, generating new revenue streams and stealing the business of these companies in the process.
As a result, most of these companies were forced to move from very permissive licenses to more closed ones preventing some types of commercial use (for the free product behind the SaaS model). Companies such as Couchbase, Cockroach Labs, Sentry, Elastic, Confluent, Redis Labs, MongoDB, Elastic, and Redis Labs all made this pivot at the risk of alienating their community and losing a lot of active members in the process.
Despite losing some users, this choice worked out well for a lot of them (and for the VCs behind the scenes), allowing them to generate much more revenue.
It marked the birth of the Fair-Code movement.
The dawn of Fair-Code
I believe that the phase we are living in right now is still mainly made of two types of projects.
On one side, we continue to see the proliferation of wholly open-source projects, especially with GitHub & GitLab allowing developers to collaborate in an easier way or HugginFace bringing together AI developers (ex: we have seen a lot of Open-Source LLM being developed by communities and labs that compete with OpenAI or Google).
But on the other side, there’s a significant drift towards adding paid features & choosing less permissive licenses from the beginning.
This evolution has been closely watched by a few venture capitalists (the minority especially in Europe )who have recognized the immense potential in marrying open-source dynamism with scalable business models.
However, the main challenge for many of these companies is preserving the Open-Source ethos while ensuring profitability. This tension led to the emergence of Fair Code which is different from Open-Source according to the Open Source Initiatives that define the boundaries of this movement.
Until 2015, most Open Source companies mostly relied on permissive licenses like the MIT license or Apache License 2.0 allowing most types of commercial usage of the underlying code.
With Fair Code, most code remains open for viewing and modification. However, licenses like Commons Clause or SSPL restrict certain commercial exploitations, ensuring the underlying companies can monetize effectively without offering their code to any potential competitor.
Most of these new companies leveraging the Fair Code ethos also have more than 90% of their code base being coded internally with external developers used more as evangelists and QA engineers. You can see in the graph below, how it affects the % of contributions made by outside developers. And the ratio of forks (number of people saving the code, often in order to contribute) to stars (it’s the “like” system of Github).
Most of the companies on the left of the graph are pretty young and have a Fair Code approach while most projects on the right are completely open projects or are older companies.
For VCs, investing in Open-Source companies has always been quite different from any other type of investment. You have to change your analysis grid and think more about community growth than revenue growth at the beginning.
In an article written in 2020, Bessemer gave a great framework for analyzing Open-Source companies stating that the more contributors you have, the better.
But with the emergence of the Fair Code approach (where contributors are less important), it’s becoming less and less true and VCs rather look at the size and engagement of the community, the speed of development, etc.
Some companies are impressive regarding these metrics while still having most of their code produced internally like Supabase which has on average more than 280 commits per week in their Github (while most good projects count between 10 and 50 commits per week).
I strongly believe that Open Source is now becoming the best way to build great businesses in numerous verticals like developer tools, no-code apps, cybersecurity, etc.
An Open Source founder I met told me :
“We chose to Open-Source the product for two reasons :
1. Most companies wouldn’t do business with us if we were not Open Source because they want to be able to have access to the code in case we are not here anymore and crash in a year + they have security issues
2. We want to leverage the community that could build connectors between our products and others.”
The Road Ahead
In the next phase, I predict that 2 types of projects will continue to thrive:
- Full Open-Source projects that might not even become companies. They could take the form of foundations financed by labs, largetech companies, etc. With the growth of AI and the emergence of LLMs, we have seen a lot of new ambitious Open-Source projects being built without even searching to monetize. I predict this category will continue to grow, especially in fields that are “critical for the future of humanity”.
- Fair Code will become more and more common, with the majority of what we previously called “Open Source Businesses” having much less permissive licenses, much more paying features, etc. These businesses will have the majority of their code readable and communities helping in the roadmap and the QA process, but they will get close to freemium community-driven businesses from all other standing points.
In Conclusion
From the rebellion of the first days to one of the best ways to create software businesses, Open Source has evolved a lot in the past year. While it used to be a movement to prevent big corporations from controlling the computer science world, it is now leveraged, and encouraged by most of the biggest companies in the world.
While some may think Open Source is taking the wrong direction, I think it’s the emergence of these new business models that will allow the overall movement to grow. Having long lacked sustainable means of monetization, Open Source wasn’t taken seriously, and a lot of companies couldn’t sustain their product development.
But now everything is in place to see this movement continue to grow exponentially. Databricks is poised to announce its IPO and might steal Red Hat’s seat as the largest Open-Source exit ever.
Many Open-Source companies are also emerging as alternatives to great SaaS businesses paving the way for a wave of SaaS 2.0 businesses leveraging the strengths of the Open-Source community
I’ll publish an article on this particular subject in the coming weeks, so be sure to follow me to stay informed about it.
If you liked the article, are an entrepreneur evolving in this industry, or just want to discuss Open Source in general, don’t hesitate to contact me at abel@redriverwest.com or Olivier who helped me to write the article at olivier@redriverwest.com
