Epi 19: Build Customer Data Pipelines to Activate Sales Opportunities – Soumyadeb Mitra, CEO of RudderStack

JC: Welcome, everybody, to another episode of the Future of BizTech. I’m your host, JC Granger and I’m here with Soumya Mitra, who is the founder and CEO of RudderStack. Soumya, thank you so much for coming on the show. Tell the audience a little bit about yourself and the company.

Soumyadeb: Thanks for having me, firstly. My name is Soumya Mitra or Soumyadeb Mitra, that’s my full name. This is my third company. I’ve been an entrepreneur all my life. I sold my previous company to a company called the 8×8, spent a year there, and then we started RudderStack in June 2019. I have a background in engineering. I worked in databases, worked in data for pretty much all my life. After my Ph.D., I joined a company called Data Domain, spent a couple of years there, and learned a lot about how to do a startup. The company went IPO, eventually was acquired so that was a great experience learning in a fast-growing startup. And then, I started a company, sold that too, did that for five years, I was the co-founder, CTO, sold that to 8×8 and then spent a year there and then started RudderStack. That’s my quick background.

JC: What about RudderStack? What kind of problems does it solve for its clients? I was taking a look at the website earlier, but why don’t you kind of give the audience the full rundown of RudderStack and exactly what it does for companies?

Soumyadeb: Yeah. We call ourselves a customer data pipeline company. What that means is, if you are any kind of a consumer or business, consumer company, or a B2B company, you want to understand what your customers are doing in your app. Let’s say if you are a FinTech company or if you are a consumer shopping company, you want to track everything that your customers are doing, all the products that they are making in the app, all the buttons they are clicking and all this stuff, right?

Soumyadeb: You may ask why do you want to do that? The first reason is to just understand your customers better. You want to understand what parts of the app they go to and what buttons they click on, what parts of the app is working better or not, and so on. That’s your product use cases.

Soumyadeb: Then, you also want to use that data for marketing. For example, this one, your prior product searches, you want to make them offers. And then, when somebody came to the checkout page and they drop off, then, you probably want to give them a discount coupon. Those kinds of marketing things, you want to do based on their prior activity in the app.

Soumyadeb: And then, there are all kinds of advanced machine learning use cases that have been based on all this data. You want to train the recommendation engine, people who click on this also click on that. How is that? Because you are tracking what people are doing in the app. Amazon does that and the company want to do that. And then, again, these are consumer, B2C kind of company use cases. They’re B2B use cases also. When you’re selling to other businesses, that’s where also you want to understand what your potential leads are doing with the computer website, what pages they go to, are they checking on the pricing page, what things are they doing in the free tier of the product, all these things you want to track so that you can sell better and you can do more personalized selling.

Soumyadeb: Now, all this requires you to collect all that data and to be able to track all that data, now, prior to, there was some solutions to do that. You can use a cloud analytics product, you can send everything to Google Analytics and analyze on that, but there has been a recent trend and all companies sprang to take more ownership of that data and sending to a third-party service, I want to collect all that myself, dump it into some kind of a data warehouse and do my own analysis, not just for data privacy and security reasons, but also generally, if you have access to that data, you can do a lot more things. You can do more analysis, more advanced analytics and so on, which you cannot do on a cloud.

Soumyadeb: Again, long way to, that’s what we enable customers to do so they can use our product and we have a hosted service, we are an open-source product, they can also download our software and run it themselves to help collect all this data.

JC: I like that it lets them keep a hold in their own data because you’re right, that has been an issue with third parties managing data and whatnot, so the fact that it allows a company to really keep their data in the house, in that sense, it’s really nice. Now, that being said, what kind of companies are typically the ones that hire you? Are you going after small businesses, mid-market, Fortune 1000, what kind of companies are typically the ones that hire you guys or use your services?

Soumyadeb: Yeah, we have a product and what we sell today is a cloud-hosted service that people can use. And then, we also have an open-source product that anyone can download and set it up. And then, we have an enterprise version of the product, which appeals to larger enterprises because there is more advanced features. We have three product offerings. Open source, anyone can set up and download. And on that, we have hundreds of companies deploying that all the way from small companies to pretty large brands like Goop, for example, is deploying that and so on. That’s on the open-source version.

Soumyadeb: On the cloud version, again, we have a spectrum of companies all the way from small startups like 10, 15 people startups to larger enterprises like Econs for example, is a fintech company there, 500 people there using RudderStack. We have some pretty large consumer brands using RudderStack, like, for example, the largest grocery delivery company out of India, out of Russia out, of Turkey. For some reason, we have good traction in the grocery delivery space, so they are all using RudderStack now.

JC: Ah. Okay.

Soumyadeb: And then, we have some large enterprises using the enterprise version of the product. There are a couple of PLCs we are doing and I don’t think I can name those companies until we close them, but these are really Fortune 100 companies who value the data ownership quite a bit and they cannot use a cloud-hosted service so they’re using RudderStack. It has been all over the place from the small companies, open-source developers to all the way to very large enterprises.

JC: Let me ask you a question, you said this was your third company, correct? This is your third business.

Soumyadeb: Yeah. Mm-hmm

JC: What motivated you to start this one? Why was RudderStack the next logical step for you after the last two companies?

Soumyadeb: Yeah. It kind of all adds up. My previous company was in the marketing tech space, so we are building a marketing tech platform for B2B businesses. Think of it as the VP of marketing at Cisco and you sell routers and you want to reach out to all IT people in Fortune 500 companies, right? How do you do that? We had a database of all that. We could show them ads, we could show them ads on Facebook, Twitter display, we could send them emails, a really complete marketing platform to do that.

Soumyadeb: I’ve been working in the marketing space and while working there, I saw that one big gap was, all these companies are trying to collect data about their users, showing how many people are investing, but they also want to understand what our existing people are doing with the product and so on. I saw the problem in my previous startup too. Now, there was this broad area called customer data platform that was taking off at the same time. I’ve been falling that space. Then I went to 8×8, we sold the company to 8×8, spent a year at 8×8 trying to solve similar problems. 8×8 was a public company, 25-year-old company, a lot of, a hundred thousand customers, over a billion people using our phone systems and we had a lot of data about those and if you are not collecting, you’re not analyzing all the data.

Soumyadeb: Then, again, all these interesting use cases we could have derived from collecting that data, as we are saying, the product analytics, marketing and so on, but there was no easy way to collect that data. There were a couple of solutions, but we had the same set of challenges, that being regulators in this field, the compliance requirements and all that stuff. There was not a great solution to help collect all the data.

Soumyadeb: That was also the impetus of it, it has problems, then I’m sure that dozens of other companies would need a solution that. Again, I think it’s not an idea that you get overnight. I’ve worked in the space for the last seven, eight years, first time faced the problem in a large public company and that was the motivation for starting RudderStack. I think most B2B companies start that way. There’s a problem in a real domain and then go in and try to solve that.

JC: What’s the user interface like once you have this data we’re so used to our data go to third parties, but that’s because they typically manage it well visually for us, make it easy to do something with. I guess my question is, how much effort have you guys put into the user experience? So even if I have all my data, how am I seeing it? Can I still API it out to a third-party if I want to, for example, for another company that maybe does really good at sorting that data in a certain way? Do I still have the option of releasing that data securely to a third party if I want to?

Soumyadeb: That’s a great question and that’s exactly what we do. There are a bunch of solutions we can help you collect all the data and dump into some kind of a data warehouse or a data lake that we dump all that data, but your insight point, it’s really hard to visualize that data. Your marketing team does not want to look into a data where it holds run sequel commands and so on. They want to manage it in a better product where it can create audiences, run campaigns. They want a complete SAS solution product, so their marketing team needs that. Your product analytics team may want to use some other cloud-hosted analytics solution for their use case.

Soumyadeb: At the same time, your data engineering team or your data science team, they don’t care about the UI. They want the raw dump into some kind of a data warehouse or data lake where they can do their machine learning training and so on. Th want the raw data, they don’t care about the UI.

Soumyadeb: You have all of these different consumers of that data in the organization, from the marketing to sales to support. They all have the best of great tools to want that data in. And that’s what we do at RudderStack. We’ll help you collect all that data and you can define policies, like this set of data should work with a product analytics solution. It can do things all like all the PII should be scrubbed off. If you are sending to, the product analytics does look at funnels and so on. You don’t need to send PII data there so you can do that through RudderStack and you can say that only this set of events go through  product analytics. You may not care about tracking every click and every product-

JC: Then, you can filter and segment who sees what in that data?

Soumyadeb: Exactly.

JC: That’s good. Okay, cool. As a marketing guy, my brain always goes there and I know that the audience can benefit from hearing how different business owners are getting their own business out there. What kind of marketing are you guys using to tell people about RudderStack and then, what have you found that’s been working the best?

Soumyadeb: Yeah. That’s a really interesting question. We are a one-and-a-half-year-old company, so by no means, we have figured it out, what is working, what is not. We’re still learning that.

JC: I’ve been doing this 20 years and I can barely figure it out. Jesus, it changes so quick.

Soumyadeb: Yeah. Exactly. But I think the first thing on that is, what person are we going to, like we are primarily selling to engineers, people and engineers both in small companies as well as large enterprises. It’s a really important selling motion and most open-source companies are like that. Some engineer in some company downloads thinks it’s working and then they talk to us for more advanced features.

Soumyadeb: The marketing, for us, is to figure out how to go and market to engineers. That’s always tricky. Engineers hate talking to salespeople. They hate marketing, the old-school marketing. They never click on ads. Most of them would have some kind of ad blocker in place. The way to reach them is to write great content. Content could be blogs, could be even code. We are an open-source company, so we have a detailed report with all the code open source.

Soumyadeb: And then, the code is also some kind of content or you’re writing some interesting blogs and then you try to promote that in communities where engineers are. Hacker News is a great community. I don’t know if you know Hacker News, but it’s one of the number one, it’s kind of like Reddit for developers. Every developer is on Hacker News and if you somehow get on the first page of Hacker News, then one-third of the world’s developers get to know about you or see you. We’ve been fortunate to get there a couple of times, both on our GitHub report, where we host our code, that went on first page. One or two of our blogs went to first page. That helped get exposure to the developer community.

Soumyadeb: And then, we have done things around Reddit and blogs, there are a couple of developer-focused blogs, like New Stack, TechCrunch, for example. Again, the way to market to developers is to write great content and somehow promote it in the communities they are on. So still figuring out, I’m not a marketer, so we are learning and seeing …

JC: That’s still helpful, though, because some people listening are, this is a very tech community for our audience. They probably have businesses that have a difficult demographic to go after also. Your advice being very content-heavy expertise style, that makes sense, so thank you for sharing that. That’s good.

JC: Where do you see the future of that cloud stack technology industry going? As an industry as a whole, where do you see your industry going in four, five years? What’s the future of that? What can people look forward to, do you think?

Soumyadeb: Yeah. I think, and this is a very interesting question and I have my own belief that may or may not align with what other people believe. If you look at the 90s, people were setting up their own stack. You’d buy Oracle and you buy SAP and then you buy Oracle database, you buy some application software like SAP and then you build your own application completely from scratch on top of that. And then that is your CRM. That’s where you track all your customer data and that is your ERP. That was the nineties.

Soumyadeb: Then came all the SAS revolution, starting with Salesforce. Everything, you don’t have to manage all that, give everything to a cloud provider. Send everything to Salesforce and they’ll track all their customer data and everything on customer data.

Soumyadeb: And then there’s a suite of SAS companies. Salesforce is tracking your customer data, your Google Analytics has your analytics data and then so on and they’re like 50 SAS companies people are putting together. Your marketing stack has started now becoming very complicated with all these SAS products for different things.

Soumyadeb: Now, the challenge with that approach is, now you start losing some control over your data. Your data is all over the place, you don’t have any consistency and then you have all these things running, like VOC marketing stacks where your data is in some kind of a marketing automation system that is linking to Salesforce, that is linking to something else. You have these one-off syncs running between 50 different places, right?

JC: Yeah.

Soumyadeb: Plus, there are challenges with data now, more and more companies are realizing that you want to take ownership of your data for various reasons like privacy, security is one thing, but also the kind of flexibility that you have if you have your own data. Salesforce only lets you access your CRM in certain ways. You can use their UI, you can use certain applications. I cannot send ClickStream data to Salesforce and use it for my training machine learning things. I want that data in somewhere else.

Soumyadeb: I feel there will be this decoupling of data and application going forward. Everything was coupled on your data. 90s was, your everything, your application and data, all in your data centers. Everything went to the cloud where your data is on the cloud, hosted and your application is also on the cloud, hosted by Salesforce or somebody else. We’ll probably see a cycle back from that, where your data would be under your control, where you have more direct access to your data.

Soumyadeb: And then, there are all kinds of applications on top. There will be a CRM application and so on, which may be cloud-hosted, managed by somebody, but the data stays with you. I believe there will be this unbundling of data and applications in maybe not five years, but definitely in 10 years and the infrastructure required to manage this data is also becoming easier.

Soumyadeb: If you look at Snowflake, it does the biggest IPO ever and why is that? More and more companies are putting all their data into Snowflake, in a cloud data warehouse so keeping track of your data is getting cheaper and easier and so on. That will only drive more of this. That’s the core thesis of cases of RudderStack. We believe once you have this decoupling of applications and your data, then you need to run these pipes. You have to move data around and so on. That’s where we play, if that makes sense.

JC: Where do you see your company in five years? Do you guys have any new features or tech that you’re releasing soon that you want to give the audience a little preview about or anything like that? Where’s RudderStack going?

Soumyadeb: Yeah. There is a three-month story to that, there’s a six-month story and there is a five-year story to that. Actually, there is a three-week study also. In two weeks, we are launching a big feature. It’s kind of details, I don’t know if it’s relevant for your audience, but generally, so far, we have been collecting, helping our customers collect ClickStream data. You have your app, you embed our SDK and you collect all the clicks and everything that happens in your app, you can track all that and we’ll collect all that, we’ll dump into your data warehouse, we’ll send it to wherever you want to send it to it. That was the piece that we had so far.

Soumyadeb: Now, one other big source of customer data is your cloud data. You have data in Salesforce, you have data in Marketo and all other cloud applications. When you try to bring that together, you have to pull data from the cloud sources also. That’s one big feature we are launching in two weeks.

Soumyadeb: The third thing we are also seeing is, okay, you got all this data into your data warehouse. And then, you did some modeling on top of the data. For example, you’ve computed a next best product to show to a customer or you computed a score for that customer. Is it a high-value customer or a low value customer? You did some competition on your warehouse, it could be a machine learning model, it could be a simple competition and you want to take that data and you want to sync it to some kind of a marketing tool.

Soumyadeb: You want to send it to some kind of an emailing tool so that you send an email based on those competitions. For high-value customers, you may want to give them some special coupons. You need that reverse pipe, take data from your warehouse back into all these tools so that you can run those campaigns. That’s the third part we are launching.

Soumyadeb: The way I see about RudderStack is even stream to pull all your ClickStream data, the cloud connectors, cloud extract, which we are calling to pull data from all your cloud applications. And then, once you put it in your warehouse, you do something, some magic there, and then you want to push it back to all these applications, so that’s the third piece we are calling the red house actions. These are the three pieces. They kind of complete the cloud, the customer data pipeline story. We are very excited to launch that in two weeks.

JC: That’s really exciting and by the time this airs, it’ll already be launched, just the delay in publishing. Do you only work with data in apps or is it web apps, mobile apps? What about just off my website, for example? I guess my last question about your company is, where specifically, or where do you mostly find your clients are using your stack? Is it mobile app companies, web app, or just e-commerce where they have a lot of data coming in from sales and stuff, things that and a lot of action on their website?

Soumyadeb: Yes. It is mobile apps and web apps. I think that’s the biggest thing. Between those two, if I had to take a guess, in the customer base, pretty much everyone has web property on mobile property. All consumer companies have that, so I think across our customers, 80% are using RudderStack to do both mobile than web tracking. But in terms of even volume, if I had to take a guess again, I think 80%, 70% would be mobile traffic because there’s just too much activity compared to web. That’s what it’s primarily used for. RudderStack is also used for service site tracking when you have a backend application and those are generating events. We can also generate events from that. We have SDKs for pretty much all the platforms all the way from the mobile apps web apps to service site SDKs.

JC: Soumya, you’ve been, like you said, an entrepreneur your whole life, it seems like, professional life anyway. If you had to give one piece of advice to the audience based on your experience, either previous or with this company, what’s the one thing that you would advise everyone, from everything you’ve learned?

Soumyadeb: Yeah. I can only give advice on being an entrepreneur and by no means I’m a super successful entrepreneur, so take that with a grain of salt, but one thing is, the market always wins. That’s one thing I’ve learned again and again. You may have the greatest product and the greatest team, but if the market is not there for what you’re building, then market wins, you lose. If you have a lousy product and a lousy team, but a great market, even then, the market wins. You’ll do okay.

JC: You just skate by, right?

Soumyadeb: Exactly. And if you have a great team and great product and a great market, that’s when magic happens. This is actually a quote from somebody. I forgot who, but the market always wins.

JC: I like that. I haven’t heard that, but I actually really like that. As a marketer, my mind takes that and the way I would see that for myself is, that just means that one of the biggest parts of the equation is targeting the right people because if the market always wins, that means that you have to make sure that you’re targeting the right market. If you pick the market first, in a way, and then reverse engineer what to provide them within, of course, something that you’re passionate about, educated about, you never just chase the money, that’s a losing equation, but if you say, “I want to help out these people,” and you say, “Well, what do I know, do I like doing that could do that?”, a you know, there’s a market for it and that seems essentially what you’re saying, with the market always wins. That’s the best equation.

Soumyadeb: I’ve learned that the hard way in my previous two companies. The first company was right out of college. I was fresh out of college, I had no idea about marketing and anything. I was engineer and totally, okay, I had this cool project I did in college. Why don’t I let go and sell this and it didn’t go anywhere, there was no market for that. Even in my previous company, we thought we had a good market when we were trying to build an ad tech platform, but the challenge then was, there was a market, but the market was very hard to penetrate. It was dominated by the big guys. The LinkedIns and the Facebooks and so on so even there, it was very hard for us to go and sell them in the market although we had a great team.

Soumyadeb: We are PhDs from Stanford and both my co-founders were, ex-McKinsey. We had the best possible team that could be on paper, but even then, it was not the greatest outcome for a company. I’m very bullish about my current team. We have a really great team and so on, but I think what is really different here is the market. It has been, again, a year and a half, it’s a long way to go, but I can see a stark difference in the first year of my previous two startups and my current startup so yeah, I think, hopefully something comes out.

JC: Well, congratulations for this next one here. Soumyadeb, thank you so much for being on the show here. How can the audience reach you or the company if they’re interested in your services or reaching out to you?

Soumyadeb: Yeah. The best way to reach me is by email. I’m kind of old school, I prefer email over social. It’s my first name, which is hard to pronounce, but yeah. It’s at [email protected] or @gmail.com, either way. That’s the best way-

JC: And then RudderStack, right? R-U-D-D-E-R-stack.com?

Soumyadeb: Yeah.

JC: Perfect. All right, listen, Soumya, thank you so much for being on the show. Really appreciate it and we’ll talk to you soon, sir.

Soumyadeb: Yeah. I’m very excited to be here. Thanks for having me again and yeah, nice to meet you.

JC: Absolutely.

Soumyadeb: All right.

LATEST Episodes

Subscribe now for Podcast Updates