Apcera | Interview with its Founder & CEO – Derek Collison
In San Francisco (CA), we meet Founder & CEO of Apcera, Derek Collison. Derek talks about his story how he came up with the idea and founded Apcera, how the current business model works, as well as he provides some advice for young entrepreneurs.
Martin: Today we are in San Francisco in the Apcera office. Hey, Derek. Who are you and what do you do?
Derek: I’m the Founder and CEO of Apcera. We’re a hundred and so people now but we’re still in a startup mode, so I do all different types of things.
Martin: When did you have this idea for starting Apcera, and what did you do before?
Derek: The idea for Apcera actually came around 2012, and it was an offshoot from some of the work I had done previously around platforms and platforms as a service. And what’s interesting about platforms as a service was the original premise was to try to accelerate the deployment of complex workloads. And what became very clear to me was that just speeding up the ability to deploy an app and to empower dev ops is necessary but not sufficient in the long term, and what you really needed was a trusted platform. And a trusted platform that was multi-cloud involved a lot of very, very hard problems, in our opinion, ones that we didn’t see the market willing to take on. You saw a lot of companies that would spin up and do some things in a couple of months’ time and then try to either get sold or try to push something out to the developer community where it’s like a toy or an additional tool in a toolbox that developers could use to try to hand assemble their things. And that still exists, right? We still see that and the ecosystem both embraces and then kicks those technologies out at a very fast rate these days.
But businesses actually need a platform that they can trust, that they can actually move into this next generation of computing where they can get more out of their own existing resources. They can utilize not only one but multiple public clouds. And it’s interesting that the public cloud, I think, originally started around how do we move from CAPEX to OPEX and who’s the cheapest on the OPEX, the race to zero. But what we’ve seen what the customers were engaging with now is that some of those public cloud vendors have gotten so big that actually it’s working against them and they’re nervous to put all their eggs in one basket, and so they want the ability to actually do things in a multi-cloud setup but they want to do it consistently and in a trusted fashion across clouds and with their own private resources.
And so Apcera was born out of trying to solve that problem: deploy diverse workloads, orchestrate them together (systems are becoming more complex), and then govern them all. Governance and security and policy are all these words that can be taken as a bad thing. It’s like, “Ugh,” and you see people’s shoulders shrug. And so Apcera’s vision and what we’ve driven towards was to make that as transparent as possible, drive it into the platform that IT operations actually cares about and delivers to their internal customers and make developers happy, but all doing it in a trusted way.
Martin: Derek, can you walk me through the first 12 or 18 months chronologically? When did you build the product? When did you talk to your first customers? When did you acquire them? So that I just have a vivid picture of how it went at the start.
Derek: Sure. So about March 2012, I went and did a design on my own of what I thought we would want to build. At that point in time, I started talking to VCs, the venture capital community, a lot of time seed rounds or some of the very early rounds or an investment in the founder or the founders, with a little bit of the idea. And as you go through subsequent rounds, all those rules change.
So in March of 2012, I was coming up with the idea. I spent about two weeks on it and then went to VCs, got funding in about April of 2012, put together the founding team, and we even met in June of 2012 to do the kickoff of “This is the vision. This is the general product that I think we should build.” And it was very different from what a lot of people had seen in terms of a startup, which was I said, “We will build something that might take us over a year to actually assemble.” And VCs usually don’t react well to something that they won’t even see for over a year, but part of our value proposition was that if we don’t take on the hard problems for our customers, they’re going to have to take them on, and then our value proposition goes away.
If you look at the notion of trust as delivered through a platform, which is what we actually sell, it has to take on a lot of these hard problems. You can’t keep asking the developers to understand all of the different rules, how you’re supposed to access the database, and what level of security do you need there. If something comes up, like a zero day exploit, who’s exposed? How are we exposed? And dev ops, in my opinion, has evolved in a very good way to allow developers to both innovate, develop and actually deploy into production applications at a much faster rate than they were allowed to do.
But don’t mistake that for the trust the business and the company in general has to being solely with dev and dev ops. It’s not that they’re not talented enough. It’s that they don’t have the cross functional awareness in a lot of the Global 2000 to deliver that trust factor, and it needs to be in the platform. This isn’t as very different than what happened with the operating systems in the ‘90s. So if you step back and you squint a little bit, in the ‘90s we had a very simplistic operating system, and as we exited, we had very complex systems that governed single computing resources, but they took a lot of things into the platform so that the developer didn’t have to worry about it. This is the same type of trend, it’s just for tens of thousands of computers. And multiple clouds and multiple private resources and bringing them all together under a single fabric. But it’s not unlike the general trend in the ‘90s where a tremendous amount of function and feature set was driven into the platform, the operating system for a single computer at the time. Of course, it’s actually doing the same thing, and us and the ecosystem is driving that. So usually you innovate over here and you experiment, and then as we settle on patterns and functionality and feature set that actually make sense, those ends get driven down into the platform. Does that make sense?
Martin: Yes. Derek, what do you think was the main driver for an investor to invest in the seed round? Was it only you as founder? What was the impact of your background, what you did before? What kind of confidence or so did you provide to the investor to invest in the seed round, especially given that you said, “It will take some time until we have something that we can ship”?
Derek: That’s a great question. Seed rounds and early stage rounds are mostly confidence in the founder or founders, and so my assumption is that they had confidence in me to actually deliver on some things. I’ve been very, very fortunate throughout my career, early in the ‘90s at TIBCO really defining middleware and messaging systems as a construct for building distributed systems, in the ‘90s through fin services at Wall Street, the recognition Gartner analyst level of defining categories, and I was fortunate enough to participate in a lot of that early on.
I spent six years at Google, and so really pushed on expanding out APIs to existing services inside of Google such that developers could get access to them easily, freely and with very little effort could actually incorporate these services into their own workflows. And that’s an important thing to understand of all SaaS companies are going down that path. So a SaaS company has a presence, they have some data, they start exposing some APIs, they might progress to add a scripting language or an environment, they glue these things together a little bit, and then you actually end up at the full-fledged application. I want to write a full-fledged application that consumes the data and services that you as a SaaS provider providing that have a huge amount of value for me as a business. The issue is that if I do that totally on my own, all of the effort you put into your servers are always up, they’re geo located all over the globe, and then I as a developer sign up for a single account and run my app which my business is betting in a cloud provider, without having the sophistication to match with what you’re trying to do. When that app fails, the business sees that app failing and then they look at the service that you’re providing as not doing that.
And so early on in Google we got that. And I didn’t participate directly in Google app engine but I was watching what it was trying to do and what problems it was trying to solve. And we were doing the developer APIs, and so the developer ecosystem inside of Google was kicked off by some of these efforts.
VMWare then came along and Paul Maritz said, “Hey, would you be willing to join VMWare?” And VMWare, by the way, started right next to TIBCO on Porter Drive in Palo Alto. But I had no inkling of why I would ever join VMWare. But what was presented to me was come up with an idea that moves up the stack for somebody like VMWare. And the idea was deploying applications into a cloud environment and with production quality, meaning it stays up and you don’t have to do a lot of things to do it, was really a market that wasn’t being served for the Global 2000. And so the idea going to VMWare was to solve that problem. And what happened was that that part of the problem was a huge success in terms of what myself and the team delivered. The issue was that I really quickly realized that that’s going to run out of runway and that somebody’s really nasty hard problems that have to be baked into the core operating system, so to speak, for data centers and cloud providers didn’t exist. And so that’s why in 2012, while the technology that I had worked on was taking off, I decided to leave because I really believed I could see the writing on the wall when this was going to run out of runway, and just making developers deploy things faster was not sufficient.
BUSINESS MODEL OF APCERA
Martin: If you were to rephrase the value proposition that Apcera is offering to its customers in 10 to 15 seconds, what would it be?
Derek: A trusted platform runs on multiple public clouds and your private resources, brings them all together in a single fabric, and allows you to do things both faster and with less headcount. The only thing inside of IT that’s getting more expensive is people, and so anytime you can actually repurpose or save headcount at being able to deploy and maintain the speed of innovation within a company, companies are very attracted to that, especially when trust actually involves security and policy and governance and all the stuff that they care about and know they need to care about, but being able to do that and still allow developers to actually be very, very agile and actually speed up.
Martin: Who are your customers and how do you acquire them? And especially who is making the purchase decision at your customers?
Derek: Most of the target area for us is the Global 2000. Our customers come in lots of different verticals, so telecom, fin services, media, insurance, but all of them roll up to we want to do a migration to another platform. So you can call it a cloud migration if you’re going from on prem to a cloud. But we’ve seen customers who were trying to move from VMWare to OpenStack, so it’s all private. We’ve seen true multi-cloud where they want to move to a public cloud but then they also want to tie in existing resources. And what’s interesting about the public clouds is that the race is on now for a class of services, so it’s not as much I’ll pick you over another cloud provider based on cost. They’re now looking at “Ooh, I might really want to run this application to consume that specific service which I don’t want to build on my own and that one of the top three big cloud providers invested. But then there’s something over here with another cloud provider that I want and I have to be able to take advantage of that.”
And so that presents customers with we have a cloud migration story, an app migration type of initiative, and we are quickly going to get out of control with the number of people we’re going to have to hire or train to understand we do it this way today. How do we do it in this public cloud provider, then how do we do it over there? What do we have to do that might be different? All of these cloud providers do something slightly differently. Still they’re the same workload underneath, but how you secure it, how you manage it, how you actually orchestrate it together and plug it together with other components is different from everyone.
And so customers are faced with “Wow, we’re going to have to hire a lot of people. How do we actually trust that what we do there translates to everywhere else?” And so Apcera immediately comes in and says, “Keep doing what you’re doing today and allow us to put a single fabric that actually is consistently enforced and driven from a governance and policy perspective, consistently across all environments so you don’t have to worry about it. So the ability to demonstrate getting an application on our platform is very trivial. If you invest in container techs or Docker type images, it’s for free. It already runs in our platform. And then you show the customer in another two to three minutes policy dictating where that workload can run and moving it between VMWare, OpenStack on premise, then to Amazon, to Google, to IBM’s SoftLayer to Microsoft’s Azure, all within about two minutes, with the system completely rehealing itself, the application always being available, that’s very powerful. And they immediately go, “That’s where we want to be.” And their mind isthat it’s going to take them two to three years to get there. We can demonstrate to them that we can get them there in a matter of months or even less.
And so now instead you’re looking at “I don’t have to hire a whole bunch of people to do this, and my three-year commitment to get there I can actually deliver this thing maybe in three to six months from a production grade quality standpoint internal to my business and my users.” That becomes very powerful.
Martin: Which professionals are you targeting? Are you targeting the head of dev ops or the CTO or CIO or whomever?
Derek: Mostly we actually sell to IT operations. So there’s usually a constituent inside of there that’s chief architect in platform services and those types of roles. We have had CIO types who said, “I have both. I have the dev ops and I have these IT ops, and I need to figure out something that these guys are going to deliver to this group to enable them to do what they want to do at speed but such that it were in a trusted fashion.”
It’s not as popular anymore but like shadow IT ops and stuff is still a thing with some of these larger companies that aren’t necessarily rooted in the echo chamber that is Silicon Valley. They have a migration path that they want to build, and they believe it’s a two- to three- to four-year journey, and us being able to quickly accelerate, demonstrate that we can do that, demonstrate that we’re a trusted partner for them, understanding that their IT budgets, which is their IT ops and a lot of the development and innovation piece, is growing maybe at 2% to 3% a year, yet the demands on the business are growing exponentially and what the expectation out of that group is is exponential yet the resources they have to spend is linear at best with a very, very small growth rate. And so us being able to come in and show them the speed at which they can actually get a platform and applications and services migrated in this fashion, in a trusted fashion where we can actually prove why it’s trusted, that’s been resonating extremely well.
Martin: Derek, you said that Apcera is basically a platform as a service. How is the revenue model working, and what’s driving the pricing?
Derek: So platform as a service, I guess the best way to describe, is container management, orchestration, policy, all words that you can use to describe us. I can’t tell you platform as a service itself is being redefined, whether it’s Apcera or anyone else. And the biggest thing to understand around that redefinition which we want to be part of is the developers have a preference and they want choice, and they’re willing to give that choice up for short term gains, but eventually the only choice that they care about, or the only preference and opinion that they care about is their own. And so delivering a trusted platform has to be able to enable their choice. So PaaS, as it was defined early on, was you the developer don’t have very many choices. The platform is going to do all the stuff for you, but you’ll give it up to speed up. What we’re seeing now with things like Docker and container management systems is no, the developers won’t have their own choice but we need a platform as an IT operations group that actually drives confidence that we’re doing the right thing.
We sell a managed service, so we actually bill subscription-based for the number of assets that you use, whether it be nodes or memory. It depends on the customer. And so we sell by saying, “How big of a platform do you want to set up?” regardless of where it is. So regardless of if it’s on premise or if it’s public cloud, the pricing model is the same.
ADVICE TO ENTREPRENEURS FROM DEREK COLLISON
Martin: Let’s talk about your advice to first time entrepreneurs. What advice could you provide to first time entrepreneurs so that they can learn from your learning experience?
Derek: There’s a lot of great lessons to be learned. And I do quite a bit of annual investing these days and I’m sitting on some advisory boards and I talk to a lot of young entrepreneurs. Coming from how Apcera approached the problem, it might sound interesting or counterintuitive to what we did, which was a very, very broad technology set that is addressing very, very fluid markets. And we purposely did this and we were trying to build a very large business out of that. But in general, the best advice I can say is concentrate incessantly on what makes you different, and anything that doesn’t make you different, don’t do that. Use someone else’s technology to do it, or not outsource it per se but don’t get caught up in the minutia of saying, “We want to deliver this value, and it involves all these things.” Keep wielding it out to, “This exactly is what makes us different,” and then maniacally focus on that and drive the value out of your customers.
Customer interaction and understanding what you do well, but more importantly what is the problem we’re trying to solve for these customers and are we meeting that goal? And that’s not something that you’re going to start on Day 1 and then say, “Okay, we’re good to go. We know what it is.” It’s a fluid process. You have to invest very, very early on and consistently iterate on what problems are they facing? What problems are we making easier for them to get through? Is it a bottom line thing? Is it a top line thing? Is it a speed thing? And be very, very clear on what those things are when you walk into your customers. And so even early on for entrepreneurs, the biggest advice I give is say, “Okay, well, if you want to do all of these things but I only tell you to do one, which one is it?” because I think, especially in Silicon Valley but I’ve seen this now in pockets all across the world, entrepreneurs really want to do good. They want to solve big problems, and I think that’s amazing. But getting going, what is the first thing that you solve and you do really, really well? And then you can grow from there. But if you grow and you have this massive undertaking and you’re not exactly clear on what problems it’s solving and how you fit into the market, that’s a challenge.
And again, it might sound counterintuitive because Apcera starting out, people who weren’t in the know or on the inside were like, “Wow, we didn’t hear anything from you for like a year.” And we were trying to solve some very, very hard problems, and that’s why. But also now we have a very broad technology offering and we have applicability in markets that are extremely fluid. PaaS is being redefined. Cloud management platforms are being redefined. Container management is a new market that’s emerging, even though it’s been around but now the analysts are starting to recognize it. And so making sure that you’re constantly evolving and being aware of how you fit into your customer’s problem set and what the analysts’ expectations are has to be job number one.
Martin: What are the patterns that you see on successful and not so successful entrepreneurs that you can share?
Derek: One of the ones I had a conversation just the other day, a lot of entrepreneurs that I’ve seen who have been successful moving into the entrepreneurial type of world is they’re extremely good at contributing individually. They usually are very controlling. They default to “Never mind. I’ll just do it myself.” And I’d have to be honest that I was probably that type of person still in 2010 or so. I think you have to commit to empowering the people that you bring on because the best things that you actually get done in life in terms of starting a company, you have to do it as a team, and it’s very hard for some entrepreneurs. They don’t want to give up control. They don’t want to give up investing in their people, so to speak, whether it’s equity or some other things.
And at least from an Apcera perspective, I’ve never regretted anything around really investing in the people, really pushing hard around things. Even when we started the company, I was by myself trying to get great health care benefits. We had no employees. It was just me. And someone looked at me and said, “That’s kind of foolish because it’s really expensive.” And I said, “Only until we hit eight people.” And the first eight people that I probably want to target are going to care deeply about this. You need to understand how you get the widest range of talent. And talent always has a different understanding of risk and reward. And even in Silicon Valley and in San Francisco, people have families. People want that work-life balance. And so investing in your people and making sure that if you’re successful, they’re successful, do that from Day 1.
A lot of piece is that I am not a massive fan of lawyers per se. Invest in a lawyer. The first call you should do before you hire anyone or sign anything is get a hold of them because you want everything to be done right if you want to set yourself up for success, whether that be potentially a new venture IPO or an M&A or large investments. And all of that stuff matters, and a lot of entrepreneurs are really, really good at the big picture and really, really good at details way, way down in the leads. They need to make sure they invest in the middle stuff. They don’t have to do it themselves but they need to make sure it’s being taken care of. And lawyers and health care and benefits and HR and all the stuff that you might not be thinking about, you want to get ahead of that, so invest in that earlier than you would expect.
Martin: Derek, thank you so much for your time and for sharing your knowledge.
Derek: Thank you. I appreciate it.
Martin: Sure. And next time if you’re thinking about starting a company, and you know that scaling is very important, I mean you can use capital for scaling and most importantly you have to build an organization with lots of people, and then if you want to scale via people and organization, you need to think about pushing ownership down because in the end if you want to control everything, you won’t be able to scale that much. Thank you so much. Great!
Derek: Nice to meet you.
In Redwood City (CA), we met with Rajeev Goel, Co-Founder and CEO of PubMatic, a leading marketing …