The below is a full (unedited), machine-generated transcript of a Youtube session / podcasting episode I recorded with Patrick Chase, principal at Redpoint Ventures in Q1 2020. You can view the video/listen to the podcast on Youtube, Apple Podcast, Stitcheror wherever you get your podcasts.
Trailer
Erasmus Elsner 0:00
Hi, and welcome to another episode of Sand Hill Road, the show where I talk to successful startup founders and investors about the companies that they built an invest in. And the goal, like always, is to give you a sense of what it’s like to be in their shoes, how their businesses take, and to learn from their successes and mistakes. And today, I’m super excited to be joined by Patrick Chase who is an investor at Redpoint.
Patrick Chase 0:20
Yeah, thank you so much for having me.
Erasmus Elsner 0:22
Together, we’re going to take a deep dive into the commercial open source segment, where Redpoint is a very active investor.
Patrick Chase 0:28
We like to invest in companies where the market is so big, that it wouldn’t matter if a cloud vendor entered the market.
Erasmus Elsner 0:37
Having invested in the likes of HashiCorp, Mattermost, MapR, Cockroach Labs, and most recently, Dgraph. But let’s hear it from Patrick himself and let’s jump right in.
Interview
I’m really honored to be joined by Patrick chase here from Redpoint today. He is an engineer turned investor, he started his career at LinkedIn. And after an MBA at Stanford, he started a Redpoint, I think it was one or two years ago. And it’s probably been quite the ride for you to transition from an engineering role into the investor role. I give the floor to you, Patrick, maybe walk us a little bit through your journey to the wonderful world of venture.
Patrick Chase 1:36
Sure, yeah, thank you so much for having me. As you mentioned, my professional career started in the machine learning world at LinkedIn. And I joined the team there. This was in 2013, right when machine learning was starting to become really popular. And once I learned about it, that you could get a computer to do a task without explicitly programming, I just wanted to go as deep as I could in that world, and really get close to hardcore machine learning. So that brought me to LinkedIn. Then while I was there, I worked on the feed and search. So if you’ve seen the trending news in your feed, that was one of the things that that we built, and it was really fun project touched a lot of open source technologies, both in the offline Big Data world and the online world. So we used to do Kafka Spark. And then also, LinkedIn had a proprietary system, but something similar to elastic. So I was a big user of open source technologies back then. Now, I’ve moved over to the investing side and investing in open source technologies. And the transition there was I had been an engineer for a couple years love the work. So felt really fortunate that I could work with the data team at LinkedIn, but decided that I wanted to do something closer to the entrepreneurial ecosystem. And so I was thinking about either starting my own company working at an early stage company, or working in venture capital. That was my motivation for going back to business school. And when I was in Business School, began working with the redpoint team. Well, everyone here loved the work, and so came back full time. And now I focus on Saas and infrastructure for red points. And within that spend most of my time looking at an open source and new opportunities. They’re
Erasmus Elsner 3:29
So I personally first came across Redpoint, reading the famous blog posts of Tomasz Tunguz. I hope I’m pronouncing his name correctly, it’s one of the hardest names to pronounce correctly. Yeah, I think he’s originally Czech, I think he once told Jason Calacanis. So and he’s really this Saas metrics guru for me in many ways. And, and so I always have this impression of red point being this venture fund that is very clear and open about the kind of metrics that entrepreneurs have to hit before they, they pitch them. But maybe from your first years at red point, give us a little bit of of your personal experience at Red points so far, and maybe also a bit of a high level introduction to red point itself.
Patrick Chase 4:18
For sure, so Redpoint has been around for 20 years, we’ve invested across consumer and enterprise. And we invest out of two funds right now an early stage Fund and the growth fund. And between those we invest all the way from seed to series D. So kind of the full lifecycle of companies. And I spend all my time on the early stage side SOC series A and Series B. Some of the things that I think, make us unique, as you were mentioning, I think, we like to be super transparent with founders. I think that’s kind of what you were talking about that and we think about that in a couple different ways. One, putting metrics out to the community. And tomash has done a really good job of this with his blog. But giving back to the community and giving them the metrics and the data that that we have as investors. And so they can think about how they’re tracking towards those metrics as building their company. And we like to be really transparent in the investing process, and kind of lead them through everything that we’re doing in real time, and give them a date where we’re going to be making our decision by so that we can really be super be super clear with with everything that’s happening. fundraising, fundraising is a time of a lot of uncertainty. And we think that by over communicating, you can take out most of that uncertainty. So that’s something that we really try to do, I think two other things that make us unique, we have great have great empathy for the we like to think we have great empathy for founders, everyone, everyone early on the early stage team has been a founder or an operator. And in many cases people did and multiple companies. And we think that that gives us that gives us grab that founders are on and we like to use that to build a relationship and and really be that first call for them if they something goes wrong or something, if there’s something to celebrate, but really be super close to them along the journey. I think the last thing is, we really emphasize teamwork, red points and equal partnership. So when the founder decides to partner with us, they don’t just they don’t just get the one partner that they’re working with the whole team. And that’s by design. That’s great.
Erasmus Elsner 6:35
Let’s maybe segue to a more broader discussion about the whole commercial open source segment. And we’ve seen the presentation from a16z earlier this this week from Peter Levine, the bulk of investments have occurred over the past 10 years, there isn’t just more open source investment, this investment is leading to bigger IPOs, where he basically talks about commercial open source, I think it’s the first time that I heard someone talk about like a lead funnel in many ways. Thinking of open source as a developer driven top of funnel exercise that you get the community involved, and you basically take it from there, you have different models and ways to monetize. And he suggests that you really iterate among the different business models and, and I had Sid Sibranji recently on the show, the founder of Gitlab, and he he really went through like a couple of iterations, you know, starting with, with a SaaS model, then doing a service model, then building custom features before he really landed on the on the open core model.
Erasmus Elsner 7:46
So what’s your personal thinking about the different models?
Patrick Chase 7:49
For sure, I think there’s there’s three main models, which a lot of people have discussed, but I’ll kind of define them a little bit just so we’re on the same page, and then go into more of our thinking around the models. The first model is support, this is the Red Hat model. And this is the pay some one of these that you can call them at midnight, if something makes if you’re if your company’s going down this path, because it’s harder to scale. And the gross margins aren’t as good. So we see fewer companies going towards the support model. The second model is the open core model, as you were talking about. One example of a company there from our portfolio is HashiCorp. And they’ve done a really good job. A risk with this model is figuring out what to put into the open source vs. What to put into the Enterprise Edition hashey core basically the job of defining this right from the beginning. So the way that they think about drawing the line is anything for the individual goes into the community version, and anything that you would need for an organization goes into the Enterprise Edition. And I think drawing that line early, really gave clarity to the community of what they could expect coming from HashiCorp. And so it’s allowed them to go build a very successful business without alienating the community. And so I think a key consideration with the open core model is, what do you end up putting into the orphan courses, or when putting into the open core. And then the last model is the Saas model. And this is where you charge for hosting or you charge for use for storing data. And we think that can be a very successful model as well. In terms of some of the trade offs there. I think the open core model is interesting, because it has it usually has the lowest cost associated with delivering a service. So we’ve seen we’ve seen this with like companies like mattermost and and the cost to serve a new additional client since they’re not hosting the infrastructure themselves is very low, and which can means that they can grow really quickly in terms of revenue. Name without a big cost on their side associated with that revenue. The flip side of that is on the Saas with with us you have a higher you have a higher cost associated with the revenue but but potentially better lock in because everyone’s storing their data with you. So those are those are some of the trade offs. This might be an oversimplification, it certainly certainly isn’t a lot. That makes a lot of sense. Yeah. A lot of areas. But I think if you are storing or processing data, then a SaaS model makes sense because you can charge for storing and processing that if you’re if you’re not, then I think the open core model makes sense.
Erasmus Elsner 10:37
So you would say like, the sweet spot, when you’re looking at companies at the moment is probably more in the open core and Saas side of things. With a services model being maybe like an addition for the onboarding, which I’ve seen a lot recently for the onboarding sort of services, but then the sweet spot being really the open core model. And, and the Saas if possible, if there’s data stored.
Patrick Chase 11:09
Yeah, we try to encourage people to go towards the open core and the Saas. Let’s be honest, a lot of companies do start with support. Yeah. And we think, engage the community. And you get people using your software early, because you do need those early adopters of cloud go out and build the Saas offerings. So we see some companies start with support and off along with our enterprise version, and then the VSS model over time.
Erasmus Elsner 11:36
Absolutely. So let’s talk a little bit about defensibility of business models. And something that we often see is that we have once there is an open-open source projects, we have different monetization layers, layered on top of it, and we’ve seen it with the Hadoop and MapR. Basically, having these three players Hortonworks, Cloudera, MapR around this open source project. And in terms of defensibility, this raises a couple of questions. And obviously, then you also have the cloud providers moving in. We’ve seen this with Redis and Redislabs and AWS basically moving in, and then the whole discussions around in the commons class. How do you think about defensibility of open source projects?
Patrick Chase 12:22
I think this is a really important question that has come up recently, for sure. I have Yeah, we think it’s a really important question. And it’s something that we think about with every open source investment that we make, right? How will this company respond, if a cloud vendor starts selling and selling it starts to sell a competitive product is something that we think about for all of our investments, I think the first thing we think about is market size. So we like to invest in companies where the market is so big, that it wouldn’t matter if a cloud vendor entered the market. And one, so one company that comes to mind is snowflake, right? They more than more than tripled their revenue over the from the past year. And they compete directly with redshift BigQuery offerings from all these cloud vendors. And one of the one of the reasons why they’ve been able to be so successful as the cloud warehouse data warehouse market is so enormous that you can still build a huge company, even with the cloud vendors going after it. So snowflake, snowflake is not open source, but they do compete directly with the cloud vendors. And we like to find companies that kind of have that similar aspect that I think the last example 5.3 bill Binney, and there was there was an article published recently that Amazon Amazon makes more money selling elastic, for more money at the same amount of money Stein elastic as elastic does. And so that market is so big that it can support multiple players. So I think that’s the first thing we think about. The second thing, the second thing is getting getting the core contributors from the open source project on board, because you want that expertise, you want to be the person that that startups think of first, when are companies think of first when when going to buy the product, a company that comes to mind is is cockroach DB from our portfolio. They are the team that built Spanner or the core contributors are the founders of cockroach dB. So if you’re going to go buy a cockroach, or buy a distributed SQL database, and you want to go then you want to go to cockroach, you don’t want to go to to AWS and I think and I think the last the last point in terms of defensibility is is using focus as a startup to build a better build a better experience than the cloud vendors. Focus is really the only advantage that you have as a sort of competing against the Amazon Googles of the world. I think there’s people that emphasize multi cloud, you know, vendor lock in if you’re not going with the AWS offering, I think there was I think there was one important, but when it comes down to it, people will pick the better experience. And I think as a startup, you need to go to go build that.
Erasmus Elsner 15:06
Yeah, for sure. Let’s talk a little bit about this famous Marc Andreessen quote, of software’s eating the world. I had Joseph Jacks on the show (my first guest actually) who’s investing exclusively in COSS. And he says that basically, in the future, open source software will be eating software. And what he means by this is basically that you will have open source software moving from basically the back end, all the way to CMS, CRM, and to, to ultimately consumer. And in your portfolio, you have Mattermost, which is the open source Slack. You mentioned that at LinkedIn, you worked with Kafka, which spun out of LinkedIn and became Confluent. I think LinkedIn is still actively really contributing, I think they have 30 engineers working and contributing to Kafka. And this is more like the back-end of this consumer application, but we could ultimately see open source moving really directly to consumer software side. Maybe that’s still a little bit far out in the future. But I’ll just put it to you. What do you think about this? Is this totally dystopian? Or could we see this in our lifetimes?
Patrick Chase 16:17
Yeah, I think the consumer software might be a little out there. But we are really excited about open source applications, open source at the application layer. And you talked about our investment in mattermost. That investment that investment has led us to take a more holistic and more holistic view across a lot of the enterprise verticals and try to figure out okay, where would where would an open source alternative? Make sense? And I think that is that’s something that’s that we’re really excited about? I think the I think the three things that we look for when looking at a vertical and deciding if it would make sense for an open source application, I like looking at Okay, should there be an open source Dropbox? Or should there be an open source Salesforce, right, the three the three questions that we asked ourselves, our first, is there, is there a desire for self hosting, is there a benefit for self hosting, the benefits, the benefits to the open source really are only there, if you’re self hosting, you get data control, flexibility, flexibility, the ability to add a lot of your own integrations. And these are things that we’ve seen with mattermost, there was and so if there wasn’t that desire to sell posts, then it’s probably not a great market for open source application. And the second thing, the second thing is we look for verticals where the software, the software is adopted by the IT department. And the reason and the reason for this is you want a technical buyer, because then they can go and go out, they can go out for the farm deploy the software themselves, and and they’re going to be care of integrate and data privacy, all of these things. So you really want it to be your your champion internally. And then I think, and then I think the third thing that we look at is we look for companies that are really targeted marketing, the the enterprise that are targeting large companies. The reason the reason for this is early, early stage companies or startups, typically typically are more than happy to go with Saas offering right there Look, the thing most slacks all of those goals. But once you reach a certain scale and start caring about data privacy and integrations and more customizability, then you might you might be more interested in an open source alternatives and and some of the benefits that go along with that, and the economies of scale of hosting yourself. And so and so we really look for the enterprise, enterprise target market, we’ve been looking for a bunch of different verticals, and trying to see which ones have those three aspects and might make sense for or an open source alternative?
Erasmus Elsner 18:59
Yeah, it makes a lot of sense. Maybe let’s talk a little bit about your most recent open source investment, which is a company called Dgraph, which is an open source graph database, which is obviously competing with the likes of Neo4j, which is a well funded, late stage player in the commercial open source space.
Maybe let’s take a step back first and let’s talk a little bit about why do we need a graph database instead of a normal database? I mean, the database segment has been quite active with TimescaleDB in the panel data space. But why do we need a graph database that is storing data about nodes and edges and connectivity? I think you’ve in your announcement article, you mentioned that you know the problem space from LinkedIn where obviously you have the social networks. But why is this a major problem space?
Patrick Chase 19:53
Yeah, I think I’ll start with some high level trends in in the database world, then dive into the graph database. We’re really excited about Dgraph. I think what we’re seeing happening in the database world is these kind of old school tools, for lack of a better word, are being broken up into a bunch of different pieces and a bunch of specialised systems. So you see, TimescaleDB popping up as a timescale database, you see Kafka for stream processing. And so there’s, it’s really becoming much more fragmented world, in the data landscape. And we think there’s a big opportunity for graph databases to become more important in that world. And he’s talking about kind of the ground at my background at LinkedIn. And one of the things we saw there was LinkedIn at all of these entities, right? People, companies jobs, the most important, the most important thing was not just looking at people just looking at jobs, it was how the people related to the jobs related to the companies. And know those relationships look like. And that was where all the value was, really. And so we really see relationships, as data becomes more complex as the thing that people the questions that people want to ask if data become more complex and related to relationships will become a first class citizen. And that will make storing a data, storing data in a data in a graph database, much more important than it as it has been in the past. And so that’s why we’re really excited about graph databases that are vertical. And I think you see that across almost every, almost every tech company that has reached a certain level of data sophistication, though, their own kind of graph database. And the goal with the graph is to bring that capability, not just from the top tech companies, but really to everyone who might need a graph database.
Erasmus Elsner 21:57
Yeah. So Redpoint, let the, I think it was 11.5 million Series A with the participation of the previous seed round investors. Maybe in this particular instance, how did you think about D graph compared to the other large open source graph databases, and in particularly Neo4js.
Patrick Chase 22:19
Sure, so our our thesis behind the investment, the investment in Dgraph, and in particular, really, they’re really three main things. And the first I think we were really excited about the team. Manish Jain, the CEO and founder of Dgraph, is he’s really one of the really rare founders, that unbelievable engineer, they’re amazing architects, but also has great product vision, and great people skills as well. And so when we met him, we were immediately excited about the opportunity to potentially work with him. And he also, he also has a deep expertise in the graph database world and has assembled a great team of distributed systems engineers to go after this problem. So we were really excited about working with them. I think that was the first thing. The second, the second thing is the technology. And this is moving more towards differentiate differentiation from other players and players in the market. Dgraph is distributed. And we’re big believers in distributed databases, we’ve invested we’ve invested in there just in case, we think that data volumes are only going to increase and having a distributed database that can scale to, to any of the need any of the needs of companies is really important. So that was that was something about technology that got us really excited. The second thing is, the second thing is they had amazing open source traction for vive sushi, when we we went and looked at it, I think we came across a post with a title shut up and take my money. And it was a user that was looking for, for an enterprise offering. And I think that’s the kind of product poll that you want to see from from the community when when making an investment. So I think, really strong open source traction there. The last thing that differentiates it from some of the other players is, is built with really modern technologies in mind. So it’s builds with graph qL from from the beginning, built in go. So it really fits with the other the other systems that engineers are using today who end up like this, their back end puzzle. And and so we really, we really liked that and we like investing behind the fans of GraphGL and Go, and we became excited about that.
Erasmus Elsner 24:36
Yeah, no, I absolutely get it. I’m personally a big fan of GraphQL. And when I read through your investment announcement, that’s something that really has struck a chord with me, I would say so maybe segue to the next prompt here. What is the rest of the year going to look like for you in terms of what what kind of deals are you looking at?
Patrick Chase 24:55
For sure. I think there’s we were talking about open source of the at the application layer. So for the rest of the year, one of the things that we’re doing is we’re going through different verticals, as I was mentioning, some of the verticals we’re excited about. One is security and identity. So thinking about if it would make sense to have an open source Password Manager, like an open source dash line or an open source Hakata. And, and I think I think there’s a bunch of compelling reasons why you might want to do that. And you get the advantage of transparency, so you know exactly how your data is being stored. You can, you can store it on your own infrastructure. And so you really get that control, I think that’s something that users are going to care about in the security and identity world. So that’s one we’re spending some time there. Another another area that I’m really excited about is open source workflow tools. So there’s, there’s a bunch of new tools that are popping up in this area. And we’re spending a lot of time trying to figure out who to who to track there and which players are, are, are most relevant. I think, I think anytime you have software that is connecting a bunch of things, there’s there’s an opportunity for open source and workflows like Zapier, things like that are kind of the ultimate connectors. And we’re looking, we think there could be an advantage to enterprises wanting to host that themselves, because they can connect all these different systems and build workflows from one on prem service to another one or to the cloud without without having to send all of their intermediate data to the cloud. And so that’s another vertical that that we’re spending some time looking at.
Erasmus Elsner 26:39
And in terms of the regional focus, which is typically San Francisco and Bay Area, are you also looking just at things like traction on GitHub, basically saying, well, even if this company isn’t in the Bay yet, we’re still looking at it. What’s your thinking around that?
Patrick Chase 26:59
Yeah, so we have historically spent more time in the United States. But I think one thing, one thing that’s really cool about open sources, it can it can highlight projects from all over the world. We do use GitHub as a way of sourcing those projects. So looking at engagement, their engagement in the community forums, engagement, Reddit and things like that. And I think it’s a good way to track projects. And I think it’s really cool that that someone can build something halfway around the world, and then, you know, we can go find them and potentially have the opportunity of working together. And I think that’s a really cool thing. And it’s one thing that makes me really excited to come into work, because you never know what you might uncover or a new open source project that that could pop up. And we’re always looking for those.
Erasmus Elsner 27:50
Yeah, I love that. But then when it comes to scaling, maybe just let me ask this as the last point there. Oftentimes, when you get to the scaling stage, investors are a little bit hesitant to allow remote work, even for commercial open source projects. What’s your stance on that? I mean, I’ve talked to to Gremlin recently, and I think they are 50% remote, but they’re still mostly in the Bay. Is your thinking around that the core of the founding team should be in the Bay?
Patrick Chase 28:20
We encourage companies to do whatever whatever they think is best for their for their startup and their company. And I think in a lot of cases that is remote work. So weren’t like you mentioned, we’re investors in Gremlin. We’re all we’re also investors in HashiCorp that has been distributed since day one. And so we so we really we like that model, you think it makes sense, especially given that tech talent can come from all over the world, and is really expensive, and potentially harder to retain in the Bay Area. So we really like distributed teams, it’s something that we’re seeing more and more from our portfolio.
Erasmus Elsner 28:57
Last point, a call to action. Where can people find out more about you, Patrick? And maybe if you want to plug in an event, the meetup or exciting things you’re attending for sure.
Patrick Chase 29:11
I will be at the GraphQL conference that is tomorrow. But but so maybe past by the time this is airing, but we’ll be there and the Dgraph will be there. But yeah, follow me on twitter at patrickachase. And look out for more posts from us around around open source, we’re going to try to put out some more content around open source of the application layer and benchmarking for open source companies. So if you’re interested, be sure to follow on Twitter, and amazing looking forward to it.
Erasmus Elsner 29:41
So Patrick, thank you so much for being with me here today. I think it’s super exciting what you’re working on, and I’m looking forward to follow your journey at Redpoint.
Patrick Chase 29:50
Awesome Erasmus. Thank you so much for having me.