00:00:15.799
Hey, hey there! Wow, this has been a dreamlike experience for the second day of the event. I can hardly believe that I'm actually awake; I'm not sure about you. What do you feel? I definitely want to extend my thanks to everyone who organized this and made it happen.
00:00:24.000
Today, we're going to continue dreaming, so you can relax and follow me. Let's begin with two important words.
00:00:34.879
I don’t have to convince anyone here that open source is great, right? Most people on stage today and yesterday are authors and maintainers of Rails, key components of the Rails ecosystem. So, I wonder if you agree with this statement: the web is dominating software development overall. Well, that’s not the full statement yet—hold on a second. I want you to think about it for a moment. You can share your thoughts during the break.
00:01:05.199
Now, let’s think about the life of a maintainer or an author of open source. This life can be quite chaotic because you have to maintain your day job to pay the bills and support your family. Meanwhile, there’s a constant stream of issues you must address: pull requests, questions, and other tasks. All the while, your own plans for your project may be in ruins. You simply might not have the time to work on the things that actually matter to you.
00:01:38.200
However, you might still love what you do and want to dedicate your life to building in the open. But you don’t want to be miserable. You want to be happy and successful. So this talk is for you. I’m going to strive to turn the minutes we have into a practical guide for maintainers of open source.
00:02:15.040
A bit about me: my name is Irina Nazarova, and I’m the CEO of Evil Martians, a software development consulting company. We specialize in helping startups, and I would be thrilled to work with any of you as our clients. I’m also a co-founder of AnyCable, and we’ll talk about that later. My background is in entrepreneurship and software development, which informs my perspective here.
00:02:52.640
You may recognize many of our open source projects at Evil Martians, such as Post Auto, Image Proxy, and Left Hook, among many others. You might be wondering: does all this open source work bring us business? Why else would we do it? As the CEO, I can confirm that our main source of business is indeed open source and the content we create around it. This is how clients find us.
00:03:35.480
Now, let's discuss AnyCable. AnyCable is a real-time server for Ruby on Rails Hotwire applications. It enhances connectivity, consistency, and scalability for your applications. We have numerous users of AnyCable and AnyCable Pro here today, including significant names such as ClickFunnels and Doximity. AnyCable is part of the Rails Foundation, and perhaps someone from CircleCI is here as well.
00:04:09.400
AnyCable is open source and licensed under MIT. On the other hand, AnyCable Pro is a paid on-premise product. AnyCable generates about a million dollars in consulting revenue for Evil Martians per year because this is how clients often find us. Additionally, let’s talk about Image Proxy, which is the fastest on-the-fly image optimization server used by well-known companies such as Supabase.
00:04:51.240
Image Proxy is also open source, while Image Proxy Pro is a paid on-premise product. Image Proxy actually graduated from Evil Martians and is now a standalone company. We were their first investors, which shows how we nurture talent and foster growth.
00:05:40.919
Now, let’s dream a little. We’ve discussed our current reality, but let’s imagine an ideal world for open source projects. In my ideal world, every open source project is backed by a powerful nonprofit organization that funds all its development and marketing efforts. We just got a new foundation for Rails, which is a step in this direction, but it’s not a superpower—it’s just a new organization. The reality we face is that it is quite different from this perfection.
00:06:33.759
What would make our reality closer to this ideal? For me, it’s about control—about being able to set my own direction. Even if the world isn't perfect, it's easier to accept it when you make your own decisions, rather than being at the mercy of someone who might decide to donate or not. Perhaps authors of open source feel the same way; they want to be in charge of their own destiny.
00:07:01.000
The path to this autonomy is commercialization. You can create a commercial product based on your open source project. First, think about your product and how you can advertise it. But then you have to face the nagging question: what if it doesn’t work out? Will I feel worse for trying?
00:07:34.280
Let’s help our hero, the maintainer. Imagine a fairy godmother looking just like Taylor Swift, standing in a lovely dress against a sunset. This fairy godmother is going to tell you there are just three steps to take. Are you ready to talk about them? The first step is to choose a business model.
00:08:06.599
In a big-picture view, there are three primary models you can consider: cloud subscriptions (also known as software as a service), dual licensing for on-premise products, and consulting contracts. Consider examples of successful companies; many utilize the first model. Think of companies like Elastic, Redis, Sentry, and many others. Some may use GPL licenses, but they still fit under the open source category. You might also think of GitHub and Treasure Data, which are built on open source.
00:08:46.800
The second model yields many examples in our ecosystem and community, such as U Mike Perham's Sidekiq, Bing C, Kafka, and others. None are particularly large by some standards, but Sidekiq is a successful business nonetheless. You can find instances of large businesses utilizing this model with open source built in Ruby; Github has a revenue share from on-premise licenses.
00:09:35.800
Consider companies like Rapid7, which is a data company built on open source Metasploit, or Puppet, in the DevOps space. These are successful teams who have built billion-dollar businesses by creating open source solutions.
00:10:25.240
Now, let’s talk about consulting. Notably, Red Hat earns billions through consulting work around open source. At Evil Martians, we can analyze the pros and cons of each model. One less obvious issue is with cloud subscriptions; larger clients tend to look for alternatives. On the contrary, on-premise businesses have much lower conversion rates.
00:11:19.760
Why do conversion rates tend to be lower for on-premise models? Because enterprises need to invest upfront in their infrastructure to run the software. Nonetheless, there is progress being made in this area, but it’s not as straightforward as the software as a service model.
00:12:15.360
Now that we've explored models for monetizing open source, it’s important to size the market. This entails a simple estimate—think of the potential number of people or organizations facing the problem your product addresses. This may involve companies or Ruby engineers within a certain geographical area.
00:12:55.660
From there, consider how much money these individuals or companies typically spend to solve this problem. Can it be done for free, or is the market willing to spend thousands? Multiply the number of target users by the average amount they are willing to spend to estimate your total addressable market.
00:13:47.240
Next, think about your role in this market. Can you be a niche player? This means that your market position is defended by something unique about your solution. Alternatively, could you be a major player, where your product could appeal to a wider audience. You can calculate your potential revenue by estimating your market share and applying it to your total addressable market.
00:14:35.680
This process might seem simple, but it’s important to undergo this exercise because it helps provide insight into your potential profitability. What if you eventually determine that your projected revenue is lower than your team's salaries, and you can’t sustain a standalone business? Don’t despair. It might result in a viable side project or hobby.
00:15:24.920
Many entrepreneurs have several side projects before developing something substantial. If your revenue projections appear high now, that’s fantastic! It might mean you have identified a lucrative market opportunity. However, you should be prepared for competition, as ventures in your niche may have backing from venture capital.
00:16:35.000
If your projected revenue falls in between those extremes—enough to support a few salaries—you are likely in a good spot. You can bootstrap your company from this revenue without facing fierce competition from venture-backed firms.
00:17:23.240
If you intend to build a venture-backed company, focus on validating your product hypothesis. Consider which business model to adopt, ensuring it’s not solely reliant on consulting; it needs to involve scalable products.
00:18:15.600
Most successful companies nowadays often have both on-premises and SaaS revenue streams. However, if you're bootstrapping, your model could be diverse—it may include consulting or combinations of consulting and products. Testing your ideas will help determine which model proves effective.
00:19:23.760
Bootstrapping can yield more than you expect. Many large software firms either didn’t receive funding or kept their round under $25 million at most, while still generating over a billion dollars in revenue. Consider your unique skills and recognize that excellent content creators thrive in today's open source environment.
00:20:37.760
Perhaps you are skilled at writing articles, for example. Many open source authors grasp the value of content creation. Ideally, launching a software service model first could facilitate reaching more users easily.
00:21:28.960
If you maintain strong connections through prior work in tech firms, you might consider the on-premise model to cater to large organizations. This could involve selling directly to those organizations, resulting in custom projects where you might also earn revenue through consulting.
00:22:26.240
However, remember that you won't know which model works best unless you test your assumptions. This involves creating a minimum viable product (MVP). An MVP is about more than just the product itself; you must account for the process of learning from user interactions.
00:23:50.200
Moreover, always define boundaries for your testing. For instance, if you plan to launch a pro version of your product, make a concrete plan like building something in two weeks and running tests one month later. Your goals for the MVP should be clear and personal.
00:24:32.240
As you develop and test your MVP, consider selling your product—even if only to a dedicated group of users facing specific problems. Keep your process manual at first, allowing you to keep costs down while seeking funding. Even selling ten licenses at $1,000 can validate your concept much more than merely discussing aspirations.
00:25:51.440
Now, let’s touch on usage-based pricing, which I aim to convince you is valuable for any product—even on-premise software. When considering your pricing strategy, think about the demand curve that rates how many users want your product at varied price points.
00:26:42.080
With traditional pricing systems where one fixed price exists, revenue generation becomes quite limited. Usage-based pricing creates multiple pricing points for varying levels of product value, allowing organizations to pay according to their needs.
00:27:38.960
Thus, if someone gains minimal value from your product, such as a smaller organization utilizing it sparingly, they should pay a smaller fee compared to larger organizations that fully leverage your offering.
00:28:26.080
Consulting may complement your offered products, as working on client projects provides invaluable learning opportunities, while also generating revenue. As you accumulate paid customers, be aware that your documentation now needs to evolve.
00:29:36.240
You may have had excellent documentation previously, but now, as you have paying customers, they care more about solutions to their specific problems. Recognize these new user personas and consider creating guides that address their needs.
00:30:16.360
You'll also want to share stories of how your product has helped others, as well as insights from your own experiences. As you publicize your work and communicate its benefits, your audience will grow.
00:30:56.080
In conclusion, you might worry about whether you will succeed in this journey. Succeeding overnight is unlikely, especially after introducing your initial MVPs, but remember that the connections you build and the experiences you create with your audience are invaluable.
00:31:30.240
Your first product might fail, but if subsequent products succeed, keep learning and evolving from your audience.
00:31:39.960
Finally, I encourage you all to dream bigger.