A brief history of open source: from community to commercialization


Photo by Alex Holyoake on unsplash
Welcome to the wechat subscription number of “create a story”: sinacchangshiji
Translated by Peter Levine / outside the stack
Source: Off-stack (ID: zhanwai_)
The open source movement, the open source software (OSS) movement, has created some of the most important and widely used technologies, including operating systems, web browsers and databases. Without open source software, our world will not work, at least not as well as it does now.
While open source has brought amazing technological innovation, business innovation – especially the recent rise of software-as-a-service (SaaS) – is also crucial to the success of the campaign. Moreover, because open source software can be used, modified and released free of charge by anyone, open source business needs different models and market entry strategies from other types of software companies.
Peter Levine is a developer, entrepreneur and investor who has worked in the open source field for more than 30 years. He recently delivered a speech called “open source: from community to commercialization”, which quoted his personal experience and interviews with dozens of open source experts.
Here is the speech. Peter traces the rise of open source software and provides a practical end-to-end framework for developing open source projects into successful businesses.
From a16z by Peter Levine; Jennifer Li
Open source was initially a marginal activity, but later became the center of software development. I have one of my favorite early anecdotes about open source. At that time, these software were not called “open source software”, but “free software”. This anecdote is related to a Linux / Unix command [~} $biff that will notify the user after receiving a new email. This command is named after a Berkeley Hippie’s dog, because it will run out and bark at the postman.
I like this story because the early status of open source was so marginalized that an important command was named after the developer’s dog.
I have worked in the field of open source for decades. At first, I was a developer of MIT’s Project Athena and open software foundation. Later, I became the CEO of XenSource, an open source company. In this computing environment, teachers and students of MIT can obtain relevant resources, such as files and programs, from any machine. This project has produced many widely used technologies, such as X window system, Kerberos).
In the past 10 years, I have worked on the boards of many companies involved in open source. From developers to board members, I see the development of open source and the growth of large companies based on open source projects.
I firmly believe that this is the best time to do open source business. Business innovation is critical to the open source movement, and I’m going to provide a framework today to bring open source projects to market.
The revival of open source is irresistible
The past decade has been a renaissance of open source. The chart above shows the situation in the past 30 years. During this period, about 200 companies have established with open source as the core technology. These companies have raised a total of more than 10 billion dollars of capital. In the past 10 years, the amount of investment has tended to be huge.
In fact, three-quarters of these companies were founded and 80% of their financing took place after 2005. I think we are still at the beginning of the open source Renaissance.
These investments are leading to larger initial public offerings (IPOs) and larger mergers and acquisitions (M & A).
It is worth noting that in 2008, mysql, a software company, was acquired by Sun Microsystems, an it and Internet technology company, for $1 billion (Sun Microsystems was later acquired by Oracle, an enterprise software company).
At that time, I was convinced that $1 billion represented the highest price any open source company could get, which for many years had been a ceiling level existence and was seen as an excellent way out for the software industry. For many years, the software industry has only regarded open source software as a commodity.
But look at what has happened in recent years. We have cloudera, mongodb, mulesoft, elastic, and GitHub, all of which are involved in multi billion dollar IPOs or M & as.
Of course, there is red hat, an open source solution provider. In 1999, the company went public for $3.6 billion, and this year, it was sold to information technology giant IBM for $34 billion. I’m very looking forward to whether there will be new purchase price ceiling in the future.
Open source is also expanding to more areas of software. Traditionally, the development of open source software mainly focuses on enterprise infrastructure, such as database and operating system (such as Linux and MySQL).
Due to the current revival of open source, almost all industries are actively developing open source software – financial technology, e-commerce, education and network security, etc.
So, what’s behind the revival of open source? To understand this, let me talk about the history of open source in my memory.
The history of open source: from free software to SaaS

Open source 0.0 — the era of free software
Open source began in the mid-1970s, when I was a programmer, and I called it the 0.0 era – “free software” era. Researchers and amateurs developed software, and the aim at that time was: free software.
As the Internet replaces the advanced research projects agency network (ARPANET), it is easier to collaborate and exchange code.
I remember working at MIT or the open software foundation, and I didn’t know where my salary came from. At that time, there was no concept of business model. If there was financial support behind the development of “free software”, it was the research funds allocated by universities or enterprises.
Open source 1.0 technology support and service era
In 1991, with the advent of Linux, open source became more and more important to enterprises. It is a better and faster way to develop core software technology. More and more basic open source technologies emerge at the historic moment, so open source communities and enterprises begin to try to promote commercialization.
In 1998, the term “open source” was born in the open software initiative. Around that time, the first real business model emerged: red hat, MySQL and many other companies provide paid technical support and services for free software. This is the first time that we have seen a viable economic model that can support the development of these organizations.
It is also worth noting that in terms of enterprise value, open source software companies are dwarfed by their competitors, proprietary software companies. When I compare red hat with Microsoft, MySQL with Oracle, XenSource with VMware, I find that the value of closed source companies is far greater than that of open source companies.
Industry insiders believe that open source software, as a commodity, can never compete with proprietary software companies in terms of potential economic value.
Open source 2.0: the era of SAAS and open core
By the mid-2000s, the valuations of these companies began to change. Cloud computing has opened up this area, enabling companies to run open source software as a service (SaaS).
Once suppliers can provide open-source services from the cloud, users will not know or care about whether the software is open-source or proprietary, resulting in the valuation of open-source and proprietary software companies is equal, which shows that open-source really has real economic and strategic value.
The wave of acquisitions, including Citrix’s acquisition of XenSource, my own startup (not to mention MySQL acquired by Sun Microsystems, which was then acquired by Oracle), has also made open source a key component of large technology companies.
In 2001, Microsoft CEO Steve Ballmer called Linux “cancer”, but now, even Microsoft uses open source technology in its technology stack and invests heavily in open source project development.
As a result, the next open source start-up could emerge from a large technology company, just as academic research labs produce research or developers’ garages.
The virtuous circle of open source
The rise of open source originates from the virtuous circle of technological and commercial innovation, which is highlighted by its development history. In terms of technology, open source is the best way to create software, because it accelerates product feedback and innovation, improves software reliability, expands technical support, promotes application process, and brings together technical talents. Open source is a technology driven model, and these characteristics have existed since the era of “free software”.
However, the full potential of open source can only be realized when technological innovation and commercial innovation are combined. If there is no business model, such as paid support, open core and SaaS model, there will be no open source revival.
Economic interests create a virtuous circle, or “flywheel”. The more business innovations we make, the larger the developer community, and thus stimulate more technological innovation, thereby increasing the economic incentives for open source.
At the end of the presentation, I will talk about what I see as open source 3.0 in the future, and point out some interesting innovations that are taking place in terms of technology and business.
But first, let’s talk about how to build an open source business.
Three pillars of the success of open source enterprises
The success of open source enterprises needs three pillars. They start in stages, one leading to the next. In order to develop sustainable business, mature companies need to maintain these pillars and develop in a balanced way:
1. Project community matching: at this stage, your open source project gathers a group of developers who actively participate in the work of open source code base to create a community. The matching degree can be measured by star, commit, pull request or contributor growth indicators on GitHub.
2. Product market matching: in this stage, users apply open source software. Relevant indicators are measured by download and usage.
3. Value market matching: this stage reflects the value proposition that customers are willing to pay. Relevant indicators are measured by income.
These three pillars run through the whole life cycle of the company, and each pillar has measurable indicators.
Project-Community Matching

Project community matching is the first pillar, which is related to the volume of key communities and the adhesion between the project and developers. Although the scale of the open source software community varies, the strong following power and rising popularity indicate that the open source software project can stimulate the strong interest of the developers. Relevant indicators include GitHub star rating, number of collaborators and submission requirements.
Many places can be the starting point for open source projects, including large companies or academic institutions. But it doesn’t matter where to start, it’s important to have a leader to drive the project, and that leader usually becomes the CEO of the business entity of the project.
The realization of project community matching requires high-frequency contact and continuous recognition of the developer community. The best leaders will strike a delicate balance between tolerance and arbitrariness – making clear decisions, pointing the way for the project, while ensuring that everyone’s ideas are heard and everyone’s contributions recognized.
Once this balance is achieved, the project will achieve healthy growth and attract more people to devote themselves to the development and dissemination of the project.
As investors, we are very inclined to fund the leaders of open source software projects, because they know the code base very well and are the defenders of the spirit and vision of the developer community.
Product market matching
When a project has leaders and a group of active partners, the next stage is to understand and measure the product market matching.
In this process, project leaders need to be clear: What problems should open source software help users solve? For whom? What alternatives exist in the market? Without a clear understanding of users and their use cases, projects can be pulled in different directions, leading to a loss of motivation.
With the answers to the above questions, we need to observe the organic application of the software, which is measured by the number of downloads. Product market matching degree is the forerunner of future sales activities.
Ideally, open source software users will become the top-level leaders of value-added products or services funnel. ——This will be described in detail in the market entry strategy section.
When studying product market matching, it is important to consider what form your commercial product will take and how to deliver value that makes people willing to pay for it.
Here I would like to point out a common pitfall: sometimes, an open source software product may be too good, and the product market matching degree may be very high, which makes the value market matching unnecessary, which also means that there is no natural space to drive revenue growth. Therefore, while promoting the organic application of software, you and your community should carefully consider what you may commercialize in the future.
Value market matching
The last, and often the most difficult, stage is to find the right value market and generate revenue. Although product market matching usually grows with individual users, value market matching is usually focused on departments and corporate buyers. The key to value market matching is to focus on what customers care about and are willing to pay for, not what makes money.
Generally, the key point of value market matching is not the function of the product itself, but the adaptation between the product and the user, as well as the type of value driven by the product. The value of open source software lies not only in its own functions, but also in its operation and maintenance advantages and large-scale operation services.
Therefore, when considering commercial products, some problems need to be considered include: is the product designed to solve the core commercial problems or provide clear operation and maintenance benefits? Is it difficult to copy the product? How difficult is it to find alternatives? Do team or institutional customers need large-scale operational services that have not been implemented in the original open source software?
Open source found that value market matching revolved around but was not limited to the following functions:
– Reliability & Availability & Security (RAS)
-Tools, add ons
-Higher performance
– monitoring
– Services
Choose a business model
The choice of business model depends on what value you can provide to your customers and how best to provide them. It should be noted that these business models are not the only ones, and the elements of multiple business models can be used to build mixed business models.
Technical support and service model is the business model in the era of open source 1.0. Red hat has really monopolized the market and achieved scale in this respect. If you decide to choose this model, you are likely to eventually become a red hat competitor.
Open core mode refers to adding value-added proprietary code on the basis of open-source software, which is a good business model for on premise software. If the software has very valuable components (such as security functions or code sets), and keeps these components proprietary, it can not affect the normal application of open source software, then open core mode will be a good choice.
It’s important to note that if you adopt an open core model, differences in the community may become a problem in determining the ownership of functions and code. I saw this in my own company and found it important to keep pace with the community.
The ultimate pitfall here is that if the community doesn’t recognize that you’re making money with proprietary components, they’ll push the project in different directions, or start a new project based on the same code base.

SaaS mode is to provide a full set of software hosting products. SaaS is a good choice if your value and competitive advantage lies in the operational excellence of the software itself. However, because SaaS is usually based on cloud hosting, there may be a potential risk that the public cloud will choose to use your open source code and compete.
Cloud and competitive advantage
Once the open source business reaches a certain maturity, there will be threats from the public cloud and licensing issues. Licensing is a controversial topic, and while it’s important, I see some companies spend too much time talking about licensing in the early days.
I also think we may overestimate the threat of public cloud providers. Although these vendors may host open source projects, so far, as far as I know, no open source company has been completely replaced by cloud providers.
For an open source company, the more important question is: if the code is not the company’s competitive advantage, what is it?
The answer goes back to the reason open source is so powerful: the community, and how you see development. Independent open source companies have three competitive advantages:
1. Enterprise customers do not want supplier lock-in.
2. Enterprise customers want to buy products directly from the people who write the code.
3. The professionalism of large companies is less than that of independent open source companies.
I think the combination of the above three conditions will form a truly competitive value-added condition, which is why we haven’t seen the situation that large cloud completely replaces independent open source companies.
Now that we’ve discussed the three pillars, let’s look at how to build an organization around them.
For an open source company, the more important question is: if the code is not the company’s competitive advantage, what is it? Community.
Enter the market: open source at the top of the funnel
The open source community is a developer driven activity at the top of the funnel. Setting up a company is to connect this open source at the top of the funnel with a strong, value driven commercial product.
The concept of funnel is not new, but for open source companies, its operation mechanism is different. In this section, I will show you how to integrate and organize open source work into business work at different stages of the funnel, and how to coordinate each stage with other stages.
Open source market entry strategy funnel is divided into four stages and key organizational functions.
-Community management of developers can enhance people’s awareness and interest in open source products.
-Effective product management brings a user base for free open source products.
-The development and business development of potential customers can evaluate the intention of these users to identify the value and sales opportunities of potential enterprise customers.
-Self service (bottom-up) and sales service (top-down) expand the value of paid products or services for enterprise customers
Let’s look at each stage and function in more detail.
The first stage: understanding and interest — developer community management
User registration and download can measure the product’s reputation, and working with developers to improve the product’s reputation is crucial to the success of the post funnel stage. In the early days of a company, the founders are often the first “evangelists”.
With the continuous development of the company, it is very important to have a team of developers with both technical expertise and excellent communication skills. Although such talents are rare, they are usually easy to find.
Once such people are found, hire them as developers, get them into meetings, social media, engage with the developer community, and explain to them the importance and value of the project.
At the same time, you need to align your sales efforts with the messages your developers deliver, and you don’t want your community manager to “sell.” They should strive to make people interested in the product and improve their understanding of the product. Any emphasis on sales is likely to damage their reputation in the community.
When setting up a company, you also need to decide whether it shares the same name and brand as an open source project. Both options can bring success to the company, and each has its own advantages and disadvantages.
Separate names, such as databricks, a big data commercialization company, and spark, a product of its own, prevent possible brand dilution and provide flexibility for subsequent commercialization licensing. The same name often provides companies with more impetus from open source software projects. However, if the open source community thinks that it is used by enterprises to make profits, it may lead to the Estranged.
Finally, the number of user registrations and downloads is the common standard to evaluate open source software and proprietary software. The key is not the evaluation object, but the accuracy. XenSource’s metrics used to be inaccurate because they included a large number of downloads that were started but not completed. Improving the evaluation method of user registration and interest can prevent download volume from becoming an indicator of vanity, which will not push the company into the next stage of the funnel.
Second stage: consideration – Product Management
The next step is to think. After establishing the developer community, your goal is to maximize the degree of connection between developers and users, the amount of software applications and value. The second phase of the open source funnel is usually done through product management.
Effective product management can play a number of roles to support this phase: managing product blueprints for closed and open source software; communicating decisions to developers and users; and conducting product analysis, gathering more insights about usage patterns, and forecasting sales opportunities.
Unlike proprietary software, open source enterprises usually have two product blueprints to manage. Chief executives and founders of open source software companies usually spend most of their time managing these product blueprints and their relationships with each other. I like to show on a separate page how the two are closely related and how they are connected, and what features each blueprint can provide.

The most successful companies and founders have a framework or guidelines to help them draw and communicate what should be paid and what should be free. For example, open source company planetscale is determined to keep any software that can lead to supplier lock-in open source, which can maintain the good will of their open source community and enterprise customers. At the same time, establishing a function comparison table helps customers and the community understand the different values provided by free software and paid software.
Transparency in research and development, and the inclusion of community feedback in product blueprints, are particularly important for maintaining community trust. Many successful open source companies are still active in contributing to the corresponding open source projects and often act as their leaders. For example, databricks contributes 10 times more to spark than any other company.
Back to the product itself, you should analyze the product, which can help you understand your users and predict the proportion of open source software users that can be converted to paying buyers. Once the user owns the product, the product usage analysis will help to determine the product market matching pillar through the value market matching pillar and the number of users who may develop from free use to paid users, so as to predict the sales opportunities. For example, if five out of every 100 users are always converted to paid users, you can build a financial model with 5% as an estimate.
It’s going to be a complex process, and you should experiment with product integration to determine the best line between free and paid. For many open source founders, this product trial is an endless journey, and the success of market entry depends on a rigorous product feedback cycle.
Stage 3: assessment and intention – lead generation and business development
The next stage of the funnel – Evaluation and intent – is to validate and refine these theories through lead generation and sales development. The goal of this stage is to find the path from the user to the enterprise buyer, and measure the success through the sales qualified lead (SQL) confirmed by the marketing and sales team.
The first part is push marketing. Push marketing should give priority to marketing activities for specific market segments according to the well-known mode in the “preacher book” of developers at the top of the funnel. By focusing on open source users, you can learn which jobs and departments are using the product, as well as the features they are interested in. Then, you can target engineering managers, software developers, and operation and maintenance technicians, or information technicians who think your products are valuable.
Next is sales development. Sales Development Representative (SDR) should not over promote, but should put the customer’s business success first, and need to be curious about the needs of users and the way they use products.
If your campaign develops leads, there are two main filters to determine if they are qualified: 1. What organization does the developer or user represent? 2. Do they download your products or participate in your projects for the purpose of larger corporate goals?
Stage 4: payment and expansion – internal sales and field sales
With qualified potential customers, enterprises are faced with two sales models. The first may be a bottom-up self-service mode, in which users in the enterprise apply products organically and pay for them. In general, this product is designed for individual users. The second is a more traditional, top-down sales service model, which deals at the department level or expands the use of products in enterprises.
Relevant Performance of Success and Failure
As my colleague Martin Casado pointed out in his speech “the new era of growth, sales and business to business (B2B)”, coordinating organic growth and enterprise sales can lead to several common failures of open source enterprises.
The first failure is that open source users will not be paid users. In this case, the product market matching is very high, but there is no value market matching.
The second failure is that the growth rate of open source software projects lags behind that of acquiring enterprise users. In this case, the product market matching may be poor.
Three failures are that commercial products reduce the reputation of your developer community. The reason may be that the proportion of exclusivity is too large and the degree of open source is not enough, so the open source projects gradually fail.
The top of the funnel provides the key elements for everything in the middle and later stages, so before the formal marketing and sales activities, first invest in the developer community, open source projects and users.
Never ignore these three core questions: who are users? Who are paying buyers? How do your open source software and commercial products provide value to everyone?
If you succeed, you may see a chart similar to the one above. The Y axis is revenue per user, and the X axis is time. This picture is based on my previous direct observation of the company when I was on the board of GitHub. It shows the top-down and bottom-up sales modes, both of which are reflected in revenue. Here’s to say: if your revenue line looks like a layered cake, it’s a good sign.
The orange line represents revenue from bottom-up sales, usually a separate revenue line. The second revenue line may come from department buyers. The sales mode is from top to bottom, and the internal sales mode is usually used. Another part of this layered cake is on-site sales, or direct sales, which can sell or expand the use of products throughout the enterprise.
To optimize each revenue line, we should not only focus on sales activities, but also find someone who can play a specific role and promote the work purposefully.

Finally, depending on the product, you may choose only self-service or internal sales. It doesn’t matter, it depends on the complexity of the product, the best occasion and the best way to use it. I do find that most open source companies combine top-down and bottom-up sales models, which usually start from the bottom-up model, and then build a system that can expand revenue.
Open source software 3.0 – open source software becomes a part of every software company
Software industry is sweeping the world, while open source is sweeping the software industry.
Today, almost all large technology companies, from Facebook to Google, are based on open source software. These companies are also setting up more open source projects of their own – for example, airbnb has more than 30 open source projects, while Google has more than 2000.
In the future, this virtuous circle will continue. In technology, AI, open source data and block chains are all new innovations. The next generation business model may include open-source software supporting advertising, large proprietary software enterprises supporting open-source projects, data-driven revenue and crypto tokens. Crypto token mode can realize blockchain.
I believe that the open source 3.0 era will broaden our thinking and definition of open source business. Open source no longer just means red hat, elastic, databricks and cloudera; it also means – or at least part of it – facebook, airbnb and Google, and any other company that is open source for key parts of the technology stack.
If we look at open source from this perspective, the ongoing Renaissance may just be in its infancy. The market and possibility of open source software is far greater than the current reality.
Community contributor: this guide is the real work of the open source team and draws on the wisdom from the open source community. In particular, we would like to thank Ali ghodsi (Chief Executive Officer of databricks), Maxime Beauchemin (project leader of superset), jiten Vaidya (Chief Executive Officer of planetscale), sugu sougoumarane (chief technical officer of planetscale), Marco Paladino. Kong CTO, Paul St John, WW sales, GitHub SVP, Jono bacon, community manager, Kevin Wang, fosa CEO, Stuart West, senior vice president of automatic operations, zeeshan yoonas, netlife chief risk officer, and heather ·Heather Meeker (licensed lawyer for open source software).