Interview with Rudderstack Founder & CEO Soumyadeb Mitra.
In a world that’s full of data and analytics, it’s nice for companies of all industries and sizes to simplify the things they do with that data as much as possible. Fortunately, there are startups like RudderStack that are doing just that. RudderStack is a customer data platform that simplifies integrations as much as possible. While relatively new to the market, RudderStack has already made great strides and is scaling at an impressive rate.
RudderStack founder and CEO Soumyadeb Mitra recently sat down with startup coach Roland Siebelink on the Midstage Startup Momentum Podcast. Soumyadeb shared many of the lessons he’s learned with RudderStack, as well as insight from being the founder of two previous startups:
Roland Siebelink: Hello, and welcome to the Midstage Startup Momentum Podcast. My name is Roland Siebelink, and I'm a startup coach, advisor, and everything else you might need for your startup. And today I'm so excited because I'm talking to Soumyadeb Mitra, who is the founder and CEO of the famous company RudderStack. Hello, Soumyadeb. Thank you for joining the podcast.
Soumyadeb Mitra: Hi, Roland. Really appreciate you having me on the podcast.
Roland Siebelink: Absolutely. RudderStack's been making the rounds in Silicon valley lately. So excited to have you on this podcast. Can you tell us for those who haven't heard about your company yet, what does RudderStack do? Who does it serve? And what difference does it make in the world for people?
Soumyadeb Mitra: RudderStack is a solution in this broad space called customer data platform. To explain what a CDP or customer data platform is, I'll give you a concrete example. Let's say you are a consumer company. You have a mobile app and you have a website. And you want to collect data about your customers. What are they doing in the app? What buttons they're clicking. What pages they're going to in the app or on the website. And why do you want to collect all the data? The number one use case is marketing analytics. You want to send this to something like Google Analytics and then build out dashboards and reports on and all your marketing. Then you want to collect this data for product analytics. You want to send this data to do something like Amplitude to do funnel analysis and drop-offs and so on. Then you want to send this data to a marketing automation system like Braze. There are multiple users, consumers of this data. And then as you grow, you want to collect this data into a data warehouse where you're dumping it into big queries. And then you build more advanced reports. You do machine learning and so on. Without a tool like RudderStack, what you have to do is build all these integrations from your app to your five, 10 different destinations. RudderStack simplifies that. Instead of you building those integrations, you just send it to RudderStack and RudderStack can send it to wherever you want to send that data to. Your engineers are happy. They don't have to go and build those integrations. Downstream users are happy because the product team wants a new destination to send the data. They want to move from Amplitude or they want to try a new tool. They can just enable that in RudderStack and the data moves to its new destination. Those are the kinds of use cases we can enable through RudderStack. And then there's an interesting category where warehouses themselves are becoming very powerful. People are dumping all the data into the warehouse. They're building rich machine learning models. And then they want to activate that data. You want the computer to score, to make a recommendation, and you want to send that back into a marketing system to send out emails. Again, that's the integration that you have to build otherwise. But then you can do it with RudderStack. Think of it as a tool that simplifies all of these customer data integrations.
Roland Siebelink: Okay, like a middleware between all those customer data integrations. That's perfect. Who is the key competitor building this in-house? All those custom integrations that you were mentioning?
Soumyadeb Mitra: Yeah. DIY, do it yourself. This is our number one biggest competitor because engineers do it themselves. And that is where our open source product comes in. We are also an open source company, so we tell our customers if you are to do it yourself, just use the open source and build on top. Why do you start from scratch? Beyond that, I think the biggest competitors in this space are custom CDPs like Segment and mParticle and so on. But DIY is what we see the most as we are scaling.
Roland Siebelink: Okay. I've had quite a few guests and also we are coaching some companies that have this problem that DIY is the biggest competitor. Can you delve a little bit into that? How did you come up with the idea of using open source as a bridge to pull in those do it yourselfers to still expose them to RudderStack?
Soumyadeb Mitra: Yeah, I think it boils down to what persona are you selling to? We sell to the engineering personnel, the data engineers, the data scientists, engineers, and developers. When you do that, any developer always has this instinct: "I don't want to pay for SaaS software." You don't want to buy some open source product. I have been a developer all my life. When I try to solve any problem, the first thing I look for is "Is there an open source solution for this?" And I'll download it, set it up, play with it, and be happy. When it comes to running things in production, that's when I don't want to do that. I'd rather pay for software, and open source enables you to do that. You can download, play with it, and be happy. And then once you decide that now you want to run this in production, then you can go for the cloud offering, which can take care of all the reliability and high availability and so on. Open source, for us, is a vehicle market for developers.
Roland Siebelink: Soumyadeb, can you talk a little bit about the history of RudderStack? How did you come into being as a company? Who did you found it with? And what did the early stage look like?
Soumyadeb Mitra: Yeah. It started, the company, in 2019. This is my third company. I've always been an entrepreneur all my life. When it comes to the third company, you have a network of people who can help you with funding and advice. Today, we are almost a team of 70 people. It's funny, today I was looking at our first board deck from a year and half where we had five people and five customers. And today, we have scaled much bigger than that.
Roland Siebelink: That's a huge growth in just a little bit more than 18 months, I think you said. That's awesome. Most of the people that you hired, what functions did you hire them for specifically?
Soumyadeb Mitra: The current team strength, 70% is engineering because what we're building is very engineering heavy, and is technically hard. Initially, it was 90% engineering. I had just one sales guy. Over time we expanded the team. Now we have marketing and product and sales, of course, and design and partnership and all the other functions, operations and people operations and so on.
Roland Siebelink: Okay. Very good. Typically, startups that raise a Series A are starting to really focus on their go-to-market motion. And it's good to hear that you've hired some functions to take that into account. How would you say you balance the efforts of priorities that you search for RudderStack now between more product/engineering priorities versus more go-to-market priorities.
Soumyadeb Mitra: Finally, everything is important. You have to wait on the product. You have to keep building the product. The real priority manifests itself in two ways. One is to look at budgets and how do you allocate your mind share? My own time, how do I spend between marketing and growth and sales versus product engineering? My mind share is easy. I have always been a product guy and engineering guy, so a lot of my time goes to product/engineering. I wish I could do more, but I spend a lot of time there. And then for other functions, you hire leaders. I'm not an expert in sales and marketing. You have people come in and they can manage those parts of the business. And then you have a management style where you give people their objectives and let them go and run. There is no need to micromanage; they're smart. They'll go and do their best. That lets me focus on product and engineering. And then I have to do things which only I can do like fundraise. The other part of the question is how do you allocate the budget. By no means have we perfected that because engineering is a big part. Engineering is where you spend a lot of money. At the same time, the other functions, the marketing and sales, the good thing about those functions is those scale with revenue. When your revenue is scaling, you keep investing in that. If it's not scaling, then you don't need that. It's not so hard to figure out what is the right balance. But again, I'm by no means an expert. We are still figuring out and hopefully have a bit of a solution next year.
Roland Siebelink: What do you look for in hiring those people that can bring in some of those skill sets that you may not feel you have for yourself? How - at such an early stage - are you able to bring in people that you can trust with such an important part of the business?
Soumyadeb Mitra: I think hiring is hard, for multiple reasons. Number one is the market is intense. A lot of startups, they're all trying to recruit the best people and so on. Secondly, how much can you judge in an interview? If you have a team of experts, then you can still judge. I have a sales leader, let's say. Then he can go and hire other sales people because he knows sales very well. But when you are doing the fire hiring, I have to interview and I have no idea about sales. How do I interview? How do you hedge against that? Number one, you try to form a network of advisors who can recommend candidates. This is a good guy. Hiring from a network is a very strong level. Secondly, if you cannot do that, then at least during the interview process, involve people like your investors, your advisors, have them interview. Maybe they can find a data point that you cannot.
Roland Siebelink: Since you've grown so fast, people talk a lot about the growing pains of a scaling company. I'm sure RudderStack has experienced some of these too. Where would you say you've had to develop the most attention in getting all these people to land well, be bought into what you're trying to build and ultimately become productive for RudderStack?
Soumyadeb Mitra: One thing I did not touch upon in the previous comment was this idea of culture fit. Now that I'm scaling the company, I do realize the value of culture fit. The culture fit is having fundamental agreement on some common set of values. How to treat customers. How do you treat other fellow employees? What do you value? And what is your management style? Once you have culture fit amongst the core set of employees, then everything happens automatically. For example, we got very lucky with finding people. I had a great culture fit initially. Initially, it was my own idea of what is a good fit. And now the team has scaled, they went ahead and hired people with whom they had good fit, which was naturally aligned with how I thought about the business. Culture plays a very, very important role in forming a cohesive team. I'm learning this way. Of course, sometimes you'll have people who are not a culture fit. You have to manage that. You have to either set expectations or let them go because they will not be successful in an environment like that anyway.
Roland Siebelink: Yes. What would you say is specific or maybe even interesting or rare about the RudderStack culture? What makes it different?
Soumyadeb Mitra: It's funny. We recently went through this exercise of writing it down. So far, it was very much in our minds, but now we wrote it down. And I think the top thing was customer first. Every company talks about being customer first. I have never worked in large companies, so I don't know what customer first means. But except my first job was in this company called Data Domain. And Data domain was Frank Slootman, who is now a big shot. He is the CEO of Snowflake. That was his first company. And there I saw him being really obsessed about customers. In support calls, even after we went public, Frank often used to join those calls because there were some concerns of the customer. The culture that he built was really impressive. And hopefully, that's what I'm trying to replicate at RudderStack. For example, I have a ton of calls, but there's one call that I attend every day is our customer support call, where we are looking through customer issues. And we are trying to create that culture, that customer issues are the most important ones. Beyond that, I think one thing we value a lot is trust. Trust is important for us for a couple of reasons. Firstly, we are a data pipeline company. Our customers really need to trust us that they're not losing data. Let's say there was an error, we lost some data. And he tcustomer never got to know about it. They are sending billions of events to us. The only way to win trust is to inform the customer, even if they never found out about it. That is one part of trust. The other part of trust is within the org, don't have micromanagement, don't have expense policies. Just trust your employees to do the right thing for the company and make them accountable to hold on to that trust. Trust is something I value a lot and. Empathy is one thing, it's very important. We are a group of people who should enjoy working with each other. High-performance is important. You have to give feedback. You have to let people go who are not really up to your standard, but do it with empathy. Empathy is something I really care about. Making money, all that stuff is secondary. We want to have fun doing what we are doing. Having a culture of empathy is important. We tell the team, diversity is also extremely important. We want to create an environment where people don't have to fit in. You don't have to change yourself to fit into the environment. You should be able to be yourself the way you are. That's what else I care a lot about. I don't even know what a typical CEO means, but we are geeks. Geeks often did not fit in normal environments. We don't want to create a culture where people have to change to fit in. And so these are things we care about.
Roland Siebelink: What's your vision for RudderStack? How big do you guys see it become to the degree you can share some numbers? That would be interesting, of course. But I'm particularly looking at the audience of this podcast are mostly earlier stage founders, people who are still working towards product-market fit or are just about reaching that. I want to inspire them with how big some of the founders think and how they see their vision and break it down into manageable goals over time.
Soumyadeb Mitra: In terms of the current state, I don't think I should be sharing the revenue numbers publicly. But it's about a couple of million dollars. We have a couple of thousand customers across the paid and free tier. We are definitely beyond the product-market-fit stage. Now it's all about figuring out how you scale the business. Take the revenue number and triple the revenue next year. It's just hard when we need to scale the team and so on. The vision, any founder wants to take the company public. Hopefully, we will be there. What really excites me is the market. Segment, our biggest competitor, recently got acquired by Twilio. But then their revenues were in the area of a hundred million dollars or more. And the market is enormous. If not a 100 times, it is 10 to 20 times more than that. That's the opportunity for us to go and scale. But again, it's all about execution now. The market is there for us to go and mess up.
Roland Siebelink: Okay. Yeah, I like that. Move fast and break things, as some other company used to say. You mentioned Segment, how do you feel like - still a relatively small startup like RudderStack - can compete effectively against a much bigger company. That’s had a lot of fame for its own and a lot of good characteristics and a leader in the field?
Soumyadeb Mitra: This is always hard. I think you need to have some unique product differentiation that some people really care about. It was our warehouse first architecture, which has implications on both the kinds of use cases you can build and also the costs you have to pay for a solution. Because of these unique differences, you find some initial customers who really care about you. And then your product catches up. You build those features and the product eventually becomes as good if not better than your competition. I think we are in a good spot. We often win head-on deals with our competition because the product has caught up. In terms of why am I excited. Segment recently got acquired. I'm a big fan of Segment. It does a phenomenal job of creating a new market, pitching the world that they need to do it this way, and then eventually getting acquired. I think they just got acquired too soon. They could have kept going. But it's great for us that they got acquired. And now after it gets acquired, then the management changes, then it goes in a different direction, and that is already happening. I don't want to get into those specifics, but most acquisitions go in a different direction, which leaves the market open for us. I could not have asked for a better scaling. They taught the world that this is an important problem. All the marketing was done by them. We just have to go and capture the market and do it very well.
Roland Siebelink: Excellent. It's like they've created the category for you to then launch more effective products in.
Soumyadeb Mitra: I don't know who said this but you don't want to be the first company in a market, in a space. You want to be the last company of the space. Facebook is the last social network, not the first one and Google is the last search engine. Hopefully, we'll be the last customer data infrastructure.
Roland Siebelink: I like that very much. What would you say - you already mentioned some of the luck you've had, Segment being acquired and other parts - has there also been some areas if you want to share with our audience that have been harder than you thought, that have maybe been disappointments and what did you do to overcome those?
Soumyadeb Mitra: That's a good question. I have to think through that. One thing we should have done better is hire faster. I think sometimes I was not as aggressive in hiring. For example, expanding the team, expanding the sales team, expanding the marketing team. Because you have that mindset of conserving cash. Before you have product-market fit, you're trying to conserve cash. But then you suddenly realize you have a product-market fit. You are just in a less aggressive mode for a longer time than it was required. Because cash is cheap, time is not. That is one learning I have.
Roland Siebelink: I like that saying, cash is cheap, time is not. That's something a lot of startups can learn from.
Soumyadeb Mitra: Exactly. Particularly in this market. Your time is the most precious thing for execution, capturing the market. And I don't think we ever were off too much. Of course, we could have done things slightly faster.
Roland Siebelink: Well, you still have been growing tremendously. That's a good sign. Last question, what would be some advice you would give to founders coming behind you? People who are just reaching product-market fit or are not quite there yet. What are some of the learnings that you could impart on them as new founders that they could learn from?
Soumyadeb Mitra: I'll go back to that thing. The only thing that matters is the market. Get to that answer as fast as you can. Some validations of the market size. I see a lot of really smart people working on really, really clever ideas, but really obscure ideas. As opposed to, there are so many problems which have billion-dollar markets but nobody wants to work in them. Finding that market is absolutely important. And that's one thing, both in the previous startups, we did not do fast enough. You hear the Airbnb story that you persist on it for years and years and finally it really works. But others say don't do that. Airbnb is an outlier. Try to get to an answer whether there is a market or not as soon as possible. And if not, move to a new idea. You have only so much seed money to figure out product-market fit.
Roland Siebelink: As you say, not only do you have only so much seed money, but also maybe even more importantly, only so much time until you lose courage and lose the energy to be driving this. Perfect. Okay. How can listeners to this podcast help you with RudderStack? Where do you want them to go? What do you want them to download? Who do you want introductions to?
Soumyadeb Mitra: There are multiple ways that they can help us. First, we're hiring. Anyone looking for an engineering or product or sales or design or dev. We are hiring across the board. Other than that, if this is a pain point they care about, if they heard of Segment, even if they don't understand the product and they're looking for a solution, they are welcome to try our free trial. We are going to launch a very generous trial here also. Please try our product and more than anything, the feedback is also extremely important for us.
Roland Siebelink: Excellent. Okay. And you're still in this stage where, as you mentioned, you're so customer-focused that your role personally, listening into the feedback and making sure people get a conversation going. That's awesome. Very cool. Well, thank you so much, Soumaydeb Mitra, for joining us here in the Midstage Startup Momentum Podcast. This has been an amazing interview and we really thank you for your time.
Soumyadeb Mitra: Thanks, Roland, for having me. I really enjoyed the conversation.Roland Siebelink talks all things tech startup and bring you interviews with tech cofounders across the world.