- Open source and the power of community can solve big problems
- Business models for open source are evolving quickly
- Rapid scale needs senior people early in the journey
- The missing layer of the web is real-time communications
- Big visions need to talk to tech and business communities
Setting the scene
Starting initially within the Amdocs business, hear how founder Amandine Le Pape (with her Co-Founder Matthew Hodgson) formed the vision for the missing layer of the web and created Matrix. As a decentralised and highly secure open standard, Matrix is now over 30 million users strong and is powering communications at many large public bodies and private companies. Element fuels the growth of the ecosystem by being a core contributor to Matrix and providing apps and services for Matrix. In this podcast Amandine describes how open source was pivotal in building the business, how the business model and GTM evolved and how they manage their rapid scaling. Element is on a path to be world changing and a defining standard, hence needs strategic thinking in the GTM challenges ahead.
How did you get to where you are?
There was definitely a moment when I knew I really wanted to build communication systems. Initially, I was more focused on the mobile networks. I found 4G amazing, and that’s where my passion started. I started with Alcatel starting to work on 3G networks themselves, until I ended up in a startup in Rennes, who were building mobile apps. I understand tech, I like tech, but I’m not very good at it- so I like to translate it for others. One day, one of our bosses came in and said that we needed to build a business case for a communication app that we’d be selling to telecom networks, so that they can compete with WhatsApp and Skype. It sounded exactly like the kind of thing I wanted to do, so I asked to pick it up. It turns out that the other half of this app was built by a team in London, and that’s when I met Matthew in 2012!
I ended up in Paris with Matthew, telling us all about the communication platform they were building at the time, and that’s how Amdocs Unified Communication started! It was an interesting job, but what I wanted was to work at a strategic level. I was always frustrated, because I didn’t know what was happening and what was the big picture. So I said, “Okay, I need to do an MBA so that I can get to that level, I can’t wait 10 years to get there, I need to learn so that I can go faster”. I did an MBA for 18 months, simultaneously, we were setting up the Unified Comms business unit. It was like a virtuous circle where I could take the theory from the studies that I could apply in real life. It was really, really fun!
We did that for a couple of years until one day dealing with telcos was too bad. We thought, ‘we’re not changing the world, we’re just making it worse by building yet another silo, what can we do to fix it?’ We brainstormed and at some point went to Amdocs management and said “although it’s a good business and successful, we’re running it as an incubated startup, but we’re not making an impact. We have an idea, either we leave and do something more fun, or if you’re up for it, you could fund a moonshot, which may or may not destroy the phone network from which you’re getting billions of dollars a year. But if it is successful, you will have a first class seat for it, and you can sell for billions of dollars of Matrix servers “. And we managed to convince them; having proven we were able to actually build a business which was profitable, they trusted us a bit and liked the vision. So, that’s how we ended up incubating Matrix for three years within Amdocs!
Was open source the first thought in building your vision?
We built Matrix on the experience of 15 years of building communication systems. Matthew’s team has been together building these since 2003, and we’ve built our own proprietary stack.
We’ve played with everything out there, and kept thinking about:
- Why isn’t there a very simple API, which allows you to build a communication system?
- Why don’t all these communication systems talk to one another?
- Why are we building silos?
- Why is it so problematic to just build a chat app?
It’s about the team sitting down and saying, “if we had to do it based on everything we know, the entirety of the experience we have, how would we do it?”. We looked at what went wrong with the other projects like XMPP, for example, and designed Matrix. We knew it needed to be an open standard. If you want this to work, if you wanted to actually give the data sovereignty to people and give them the choice to choose the app they actually like to communicate, give them the choice to choose who is hosting their data, it needs to be an open standard. We’re building communication systems, so you need the network effect. So for that, you need a glossy app, and that’s how we put it at the top of the list.
For Element, the idea was to build an ecosystem from the get go. You can only build an ecosystem if you go open source, if you provide open source implementations of the server, open source implementation of the clients. Trying to do open source software in a multinational corporation was a bit hard. We were just at the beginning of open source being hype. Luckily, we had the support of the management. But, it’s really fun to look at the decks we built in 2014, when we were pitching for it and see the several steps: one, it needs to be open source. Secondly, it needs to be an independent brand, completely unrelated to Amdocs because we cannot build an ecosystem. We needed our own budget and be able to do the marketing we want whenever we want. It was like getting even more into independent mode. Honestly, I’m amazed that when we actually had the support we had, we wouldn’t be here without people being a bit open minded at the higher level.
Open source business models have evolved rapidly
When we started building this, of course, we had to pitch a business model. We sat down and thought, ‘what can we build on top of this Matrix thing that we’ve done?’ We came up with 72 business models. We decided we’re going to focus on collaboration and messaging, because that’s something we knew well and it’s going to drive it pretty quickly. SaaS hosting was always our main target, until the day we had the French government knocking on the door asking for our professional services. That’s how we ended up starting mostly with level three support and consulting to help them deploy the open source software. There was not much proprietary at the time, still keeping in mind that we really wanted the SaaS platform and the App Store model- that’s been the long term goal from the beginning.
Now we’re doing more and more liscencing with very specific products, proprietary products for enterprise and different use cases, but always making sure that open source product is the state of the art for day to day communication and the mass market. We are not looking at open core models where the open source is crippled, so that people pay to get all the features, because we know that otherwise, someone else will just come for it and add whatever feature we’ve been crippling out of it. Make sure that the product is useful for everyone and has all the useful features.
Starting with early big deployments
First, we had the experience of running carrier grade deployments like this when we were selling to telcos. It was just a matter of making sure the tech was up to it. We have the big public matrix.org server, which is a very good guinea pig for many users on it. Today, it’s like 12 million users on this single server. Then when the French government came along, we started with a trial and it took a few weeks to actually start the first deployments and then one year later we actually properly launched it.
Transitioning from Founder Led selling
When we eventually managed to get to Series A money at the end of 2019, we were able to have the means to actually hire a proper GTM team- until then, we were 30 people, almost only engineers. We hired our go to market team at the beginning of 2020, almost 3 years after funding the company.
Building the TAM and ICP in open source
We always had this goal and the long term vision where we believe that an App Store for Matrix is going to be where the main value is going to come. Turns out, we had to start with the public sector, which on one hand was not necessarily what we wanted to do, but on the other hand is like big flagship customers that you can plaster all over the place and bring quite a lot of money as well, which is always useful. But we needed to get out of this. So, we had to build a plan on how to go from the public sector to a mass market where we sell widgets and integrations, and think who are the next people who are going to be interested in that- if the government likes it, who will like it?
We had to build it by looking at who would be the next audience. Also, we’ve always been getting a lot of inbound because the Matrix open source project; all the techies of the world that have known about it, and are interested in it then take it and apply it to their own use cases, they bring it into their companies. It’s a very bottom up approach where you have the champions wearing the Matrix hoodies at the back of the room when you pitch, and that’s where a lot of our initial sales were. Once we had the technical teams pushing the project, how could we amplify that? First, we needed to find a marketing person who could actually grasp the size of what we’re doing, the long term vision and understand the technical bits a bit. Then we target the C-levels by advertising the work we’ve done with the government to show what we can do. Then build the sales team and adapt it as the audience changes.
They have to understand the vision, they have to understand that it’s about the long term and not the two year flip. We cannot optimise for the short term because in the end Element is here to support Matrix and grow the ecosystem. People coming on board really need to understand open source and need to understand this idea of building an ecosystem. It’s been very hard to find salespeople, it’s taken a lot of time to find an appropriate VP Sales to join the team, but hopefully we’re good now!
The other thing we’re asking our salespeople is to be very technical. A lot of them, they may not all have a technical background, but they’re a geek at heart, because they’re selling to tech people. They really have to understand the difference of what it means to have the server deployed on premise or in the cloud and the decentralisation, the end to end encryption. That’s something which is not necessarily related to open source, but for us it’s very important.
Building a company with remote working
We are a remote company. It just happened because we’ve been hiring people from the open source community and we’re developing remote collaboration tools. We still have an office in London, and one in France, where a lot of people were quite keen to go. Matthew and I really like going to the office and being there with the team. Overall, for people who were used to going to the office, it may not have been super easy to work at home all the time during the pandemic, but in itself, it didn’t disrupt the way we work too much because it was already the case.
Working with developers
I’m very used to being in a technical world and talking to developers. I’ve always been part of the 10% women, so that’s definitely the more comfortable side of the world for me. There is a bit of adapting for me now that I have to manage a commercial team, but in the end you hire people you get on well with, people who understand tech and are a bit technical too, so it helps.
Hiring senior VPs early to scale the vision
I think the latest lightbulb moment we’ve had was a few weeks ago, was when we realised we needed to grow our management team. When we got funding from Notion, Maddy Cross, who was Head of Talent at the time, showed us a very impressive presentation, about unicorns in VC and said how unicorns hire six VPs in the first three months of funding. We weren’t super keen on hiring very expensive people who wouldn’t want to get their hands dirty, we wanted people to grow from the inside. But lately we realised we’re on the critical path of everything, and we cannot go on holiday. It’s also not sustainable. So it took us one year to realise we really needed to hire some VPs if we wanted to grow. So, I think that’s the main gotcha moment, saying, “oh, okay, that was not a smart move. But at least we’ve tried”. The interesting thing about being involved in every bit of the company is that now we understand them, now we can hire the right people, we know precisely what we need. Yes, it’s probably slowed us down having us in the critical path, but now we can fix it.
Who do you admire?
GitLab and Automattic are definitely companies we like, because they’ve been very successful whilst building on open source in a very ethical way. One of our big things, these days, is market positioning. We want to build tools which are used by professionals, but are usable for individuals. That’s when you start looking at how Apple does it; how do you manage to get all these people who know nothing about tech to buy crazy computers, which are used by designers or developers? In terms of marketing position, we look at how we can do it the same.
Don’t fake open source – give to the community
If you do open source, do it honestly- don’t fake it! That’s the only way you’re going to build a successful open source business. You have to take care of your community, you have to give to your community, you have to make sure that you make the right choices. So, you’re not trying to hinder the open source product, but make sure that it can grow and you put the power behind it in terms of evangelization and community management. There are so many companies who just go open source just for the marketing of it; like they had a proprietary product, turn it open source, and leave it sitting here. What is it bringing them aside from checking a checkbox? Once you have built an open source community, it’s going to bring you customers, it’s going to bring you visibility. If you do open source, you have to do it properly, otherwise it’s a waste of time.
Celebrate women’s success in business
I think it’s quite important to make sure that every woman feels empowered to actually jump into this and do it themselves. I’ve always been part of the 10%, and I find myself quite lucky to be here, but I think there are always opportunities for everyone. I feel as women, we often don’t celebrate enough compared to men. So, let’s make sure our successes are visible, and let’s make sure we celebrate women’s successes, but not in a way where it appears like an exception. If you show to the rest of the world that this woman is exceptional, talk about how what she did is exceptional. It’s not because she’s a woman, it’s because what she did is exceptional.