Bonus Episode! Docker Is Kind of Acquired By … Who Is Mirantis?

Jon Christensen: All right, here we are joining you live with Chris Hickman on some breaking news that we’re going to discuss. Chris, what did we have?

Chris Hickman: Well, so yeah this is a bonus episode of Mobycast, figured we would get together and have a chat because some big changes happening for Docker, the company. And yeah, they made some big announcements November 13th, and it was to some extent surprising and to some other other ways it’s not so surprising.

Jon Christensen: I think we had made the prediction on Mobycast, hadn’t we? I don’t remember which episode but we had predicted that they were going to be sold. Oh, it was probably the one after DockerCon, right?

Chris Hickman: Yeah. I mean, actually pretty interesting. I was going back through my notes. And so I like to take notes at each one of these conferences and I usually end up kind of writing some like what are the takeaways and what are some action items and predictions. I actually went back to my notes for DockerCon 2018, and for whatever reason I had a little predictions section. I mean, it might be interesting just to read some of this stuff. This is after DockerCon 2018, which was in-

Jon Christensen: Over a year?

Chris Hickman: Yeah. I wrote Swarm as a container orchestrator is going to get less attention as Kubernetes becomes more popular. Docker EE customers overwhelmingly choose Kubernetes over Swarm, and eventually Docker decides to sunset Swarm. Docker is currently acting like a company that wants to go public, but I don’t think they’ll get enough Docker EE sales to warrant an IPO. Instead, I think there’ll be acquired by a larger company with the most likely suitor being Microsoft, where enterprise is king. Note that Docker and Microsoft already have a collaborative relationship in place since Docker and Microsoft have been working together since 2014.
Those were some of the notes that I had back then, and not all the details are right, but I think the overall gist. This is not rocket science either, right? Docker has really struggled, I think, trying to find its business model. Its executive office seems like it’s had a revolving door the last few years. They’ve gone through four different CEOs since 2017 now.

Jon Christensen: That’s so many. Can you imagine being an employee there? Come on. It would just be like, “Whatever. What’s going on on the internet today? Why bother working?”

Chris Hickman: Yeah. Maybe we should just say like what happened a couple-

Jon Christensen: Yeah, yeah. Let’s do the journalistic thing, the who, what, where and why. Let’s do it.

Chris Hickman: Yeah, so November 13th, the announcements were made and basically it was two big announcements. One was Docker said, “Hey, we’ve sold off our Enterprise business. It’s been acquired by a company called Mirantis ,and we’re going to be focusing on the Developer going forward.” That was one of the big bombshells.
And then the other big bombshell was, “Oh, by the way, we have a new CEO and we just got a $35 million lifeline from some existing investors.”

Jon Christensen: Employees rolling their eyes, “I guess I have to work now.”

Chris Hickman: Think about this. Okay, they sold their Enterprise business. And so let’s just put some numbers around that. 750 Enterprise customers, which comprises a third of the Fortune 100 and 20% of the Global 500, right? That sounds really good. But at the same time, you can also look at it as they only have 750 paying customers in total, right, so there’s that.

Jon Christensen: If each one of those customers is worth more than a million dollars, then that’s a billion dollar company.

Chris Hickman: Yeah, I think I saw somewhere where it was equated to about $52 million in revenue.

Jon Christensen: Yeah, each one of those customers is not worth $1 million.

Chris Hickman: No, probably not. The other interesting thing is that this does account for about 300 out of the 400 Docker employees, so 75% of the company goes to Mirantis as part of this, okay? And the CEO of Mirantis, he has said that, “Hey, this acquisition accounts for about 90% of Dockers business.” All right, so what I’m getting to here is that Docker has now sold off 90% of its business to this other company and 300 of just top-notch experts in the cloud native space, right, are a part of that. You’ve got a third of the Fortune 100, 20% of the Global 500; and yet, they have to go get a $35 million lifeline. How much money did they get for this acquisition? If they had to go get this $35 million lifeline?

Jon Christensen: Oh yeah, that’s a good point.

Chris Hickman: Did they just give it to them, like “Please just make sure our employees have jobs.” It’s a bit of a head scratcher there. Obviously, they did not get a lot.

Jon Christensen: It is, yeah. Right. Yeah, if they got $100 million, then getting another 35 on top of that would be like, why bother? That just doesn’t make sense.

Chris Hickman: Yeah. But if they got $10 million for it, right, then hey.

Jon Christensen: Yeah. Yeah.

Chris Hickman: Who knows? But it’s really kind of strange that they had to go raise $35 million after selling off their big part of their business that’s you know…

Jon Christensen: I mean, actually that sort of mind-blowing because even on just a regular consulting business, which is what it would be, right? Here’s our Enterprise consulting business. You would generally figure that there would be a multiple on the revenue and it’s likely that there wasn’t based what you’re saying. It’s likely that it was a fraction.

Chris Hickman: Yeah, probably. And even another way to look at it is, just call this an acqui-hire. You’re getting 300 folks. I mean usually acqui-hires, you value them about a million a head, so that would be $300 million. And if they got that, then again, why would they go raise $35 million?

Jon Christensen: Right. There’s something almost fishy there, right? Because it’s like there is a working business there, there’s customers and there’s revenue getting made. Unless the argument is the only reason that there’s customers there are that they bought in when they was looking dreamy and great, and they’re all on their way out the door essentially, like there’s a bunch of contracts about to end. All this revenue that they’ve made so far is not going to be recurring. They’re having a hard time getting new projects. This the ship is sinking, help us get us out of here. Nobody, no new Docker… You see what I mean? If it was even a healthy services business gives a multiple, but an unhealthy services business might get a fraction.

Chris Hickman: Obviously, it definitely wasn’t a healthy business, otherwise they wouldn’t have sold it off because what are they left with? They have no business model now, right? They have to go create a new one.

Jon Christensen: I guess I’m maybe not being clear. I’m trying to find that, not like… Okay, so the high-flying valuations that Docker had awhile back, no services business is going to get those, right? If all you are as a services business, you got to be huge in order to get those high-flying valuations. But that doesn’t mean a services business is a bad thing. Maybe it’s not worth a couple billion dollars, but it’s still it could be a nice $100 million, $150 million services business. That’s great. Sell that.
But what I’m saying is even the services business, even the one thing that they were able to do to make money in itself, without all the rest of the horridness of too high valuations for the rest of the company that could not be realized into revenue, even just the services business alone feels to me like it wasn’t doing well. It should be able to… How hard is it to make a services business? We know a ton about Docker. We have 300 people that are really good at cloud computing. You should be able to sell that. You should be able to sell projects, right? It shouldn’t be that hard.

Chris Hickman: Yeah.

Jon Christensen: Anyway, I digress a little bit, but yeah.

Chris Hickman: And then just to be clear, when you say services, you’re not talking professional services, you’re talking about software as a service, right?

Jon Christensen: No, no. I was talking about professional services.

Chris Hickman: Professional services?

Jon Christensen: Yes, yes. That’s what I mean by a services business like selling projects, getting people in doing projects for the Enterprise. That is what I understood most of the Enterprise money to be. I guess there was also some licensing that they were doing around Docker enterprise, maybe that’s the problem. Nobody’s buying new Docker Enterprise and the only way to sell services is by history of that channel. That’s it. There’s the explanation right there. If they were just selling, “We’re really good at doing Docker projects in the cloud,” they’d probably make more money then by saying, “And you have to buy licenses to this Docker Enterprise in order to get us to do that work for you.”

Chris Hickman: Yeah. They had very little in the way of ProServe. It was all about licensing Docker Enterprise.

Jon Christensen: Oh, okay. I thought that was what it was.

Chris Hickman: No. In fact, they even-

Jon Christensen: That shows my ignorance.

Chris Hickman: I mean, there’s a lot going on. Did you remember we talked about this rapid pace of innovation and there’s just tons of things going on.

Jon Christensen: Why don’t you know what’s happening inside Docker, Jon Christensen?

Chris Hickman: I do. What’s wrong with you? Keep up, man.

Jon Christensen: All right. All right.

Chris Hickman: Yeah, so they had software licensing that was their primary business and you had the Community Edition, which becomes Docker Desktop, or you have the Enterprise Edition, and that was their whole business model they were trying to get into the enterprises. And say basically, “We are your hybrid, your Multicloud, your on-prem solution for running containers.”
The big mistake they made though is they didn’t pay enough attention to Kubernetes and it was too little too late. And they put everything on Swarm from the get-go. They built their own. We talked about this in the past. They had their own orchestration system called Swarm. They’ve stuck with it. And at the end of the day, Kubernetes has won loud and clear and it really is.
Basically using Docker, you can think of it as in two different phases, right? There’s the compile time and then there’s the runtime. The compile time, that’s the developer and how do I go about building with Docker, making my images, having them in repositories, just that whole developer experience. How do I know what technologies to use and integrate with? Where do I host these things, these images? How do I share them? How do I have public versus private ones? That’s the whole developer experience as the, the compile time experience. And then you have the runtime, which is, okay, I have my images. I now want to run my containers in production and just run my business logic on this, and what tools do I need there?
Kubernetes is definitely runtime, right? I mean, it’s all about like, how do I run my containers in production versus Docker and Docker Desktop. Those tools, it’s all about the developer. And people that use Kubernetes, they still have to use things like Docker Desktop, right, or some of those tools for just building images.

Jon Christensen: Is this just part of an overall story in the industry around runtime types of companies? If you think you’re going to compete against the public cloud operators on a runtime business, be prepared to not succeed, right? Kubernetes is not a Google product, but it came out of Google and has that sort of cachet that like, “Oh, well, Google knows how to run things at scale, so I’m going to use that.” And then we’re seeing the same thing with businesses trying to compete with AWS. We talked about DocumentDB and MongoDB last year. It’s sort of like a recurring theme, unless you’re a cloud operator, you got to have a really good story for competing in that sort of operational runtime space.

Chris Hickman: Yeah. I mean, I think at the end of the day, it’s pretty much a safe conclusion to say that the money is in the runtime and not in the compile time. I think one of the best examples of this is Microsoft, right? How did Microsoft ended up being a multi-billion dollar juggernaut? They made the best possible developer experience, the best toolkits, the best IDE, the best compilers, the best libraries, the best documentation, the best support. And a lot of that stuff, they really didn’t make a lot of money off of it, right? I mean, they sold licenses to visual studio and whatnot, but that wasn’t the bulk of their revenue.
The bulk of the revenue was on selling licenses to Windows because you needed that in order to run this stuff in production, right, and that’s where they made all their money. That’s kind of like I think the problem here is that it’s unfortunate. I mean, Docker had the developers, everyone was using it, but they just weren’t able to have that strategy in place to execute on what is the runtime model. They went and tried to do it all themselves and it was just really difficult to onboard and to start using Enterprise Edition. And as difficult and as complicated as Kubernetes is, it’s still a lot easier to onboard with that than it is with Docker Enterprise Edition.

Jon Christensen: Interesting, interesting. The public clouds are getting all the runtime money right now, just like Microsoft was getting all the runtime, money in the 90s and early 2000s. And so the only place left is the spaces in between, the stuff that the public clouds are not paying attention to or the stuff that specifically has the capability of going Multicloud.

Chris Hickman: Yeah. I think the big winner here is Mirantis. They got this for a song

Jon Christensen: And now all of a sudden everyone’s heard of them. I have definitely not heard of them before and I’m sure I’m not in the minority.

Chris Hickman: No. I had not heard of them before as well, and there’s a reason for that, right? They started off as a well-funded OpenStack distro and they’ve since kind of refined that and pivoted to… They basically two main product lines. One is their cloud platform, which is basically, how do you run Kubernetes in private clouds powered by, they have their own infrastructure as code solution. That’s the one. Basically, Kubernetes for private clouds.
And then the other product line is, they call their application platform. It’s based on Spinnaker, which is the open source project spun out from Netflix. That’s the CI/CD system. They have these product lines that are based on open source software for basically running your Kubernetes containers and then also for having a CI/CD pipeline.
Again, they’re really focused on private clouds so on-prem, and then of course by extension hybrid, and then now, Multicloud, right? That was one of the things that they’re getting with Docker. And so now they have this really good… Just the technology alone.
Prediction, the label Docker Enterprise Edition, they’re saying right now they’re going to keep it. That’s going away. I guarantee you.

Jon Christensen: I can’t imagine wanting to keep that.

Chris Hickman: I guarantee that is dead.

Jon Christensen: That’s a poisoned brand right now.

Chris Hickman: Mirantis is going to take… They say they’re getting 300 employees with this. I guarantee you not all 300 are going to sign on at all. I mean, if they get half of that, I think they’re going to be lucky unless they just throw out some really good golden handcuffs for folks, right, and lock them down. But I think the whole Docker brand will go away with them, and they’re really just going to use this technology and whatever folks stay on board to have that Multicloud experience for Kubernetes. That’s a good business.

Jon Christensen: I guess that’s something I don’t quite understand. What does Docker Enterprise do that Kubernetes doesn’t in terms of Multicloud?

Chris Hickman: Docker Enterprise Edition has just a lot in the way of ControlPlane and just UI, and the ability to provision things and applications and clusters and services, and to mix and match. Enterprise Edition would allow you to even mix and match on-prem with servers that are in public clouds and then in multiple public clouds as well, whether it be like an Azure or AWS. They have all that kind of integration code and UI and consoles and just that whole ControlPlane there to support it, right? That’s one of the big things they’re getting here with that.

Jon Christensen: Okay, great.

Chris Hickman: You can almost say they’re taking a page out of the HashiCorp’s playbook. HashiCorp, they’ve really found their mark in more Multicloud for things like secrets management with Vault or a service discovery with Console.

Jon Christensen: Or infrastructure as a service.

Chris Hickman: Yeah. Infrastructure as code with Terraform, right.

Jon Christensen: I’m sorry. Yeah, as code.

Chris Hickman: Yeah. They’re all about Multicloud and people… We’ve talked about this before on a previous podcast, right? There’s these folks that say like, “Hey, I don’t want to be locked in to a single cloud vendor, right. I want to be a Multicloud.” And we’ve kind of discussed why that’s really expensive if you do want to be a Multicloud. You got to be careful, right? And you really have to be prepared to make that investment to support that, but there are-

Jon Christensen: I’m prepared to change my mind on that right here, Chris. I think that, essentially, right now today, I’m saying that there’s nothing better for a company than just pouring all of their money and resources and funding into us. Let us do what we want. Let us build out big structures, lots of connections in between them, among clouds, tons of software that doesn’t do really anything except for talk to other software. Let’s do it. I like this vision of the future.

Chris Hickman: Okay. Just print money, huh? Yeah.

Jon Christensen: Yes.

Chris Hickman: Sure. Yeah. That stuff goes on. There are companies that do get that kind of work and are doing that. Yeah, I think Mirantis is the big winner. I think Docker is definitely struggling big time, obviously.
I think another prediction that’s pretty safe to make is, so one, they didn’t get much money for selling off their Enterprise business to Mirantis. That’s why they needed the $35 million lifeline that’s going to last them long enough for them to sell that to someone else. Docker will not exist a year from now. They will not exist as an independent company.

Jon Christensen: Yeah. If they do exist a year from now as an independent company, well gosh, it’s like they will have failed again, right? Because they’ll be running out of money, running out of that $35 million. I mean, somebody’s got to pick them up, right? They’ve got to get through this phase and have somebody pick them up before they run out of that cash because I can’t think of a credible story for them to be able to get cashflow positive between now and when they run out of that cash. And if they still have a hundred people working there, that cash is going to last a year.

Chris Hickman: Yep. What’s left after this, right? They’re left with… They have Docker Desktop and they have Docker Hub. Those are actually two big crown jewels that they get to keep. We’ve already talked about how everyone that is working with containers almost… It’s something like 96% of people working with containers are using Docker and Docker Desktop, right? That developer experience, it’s still there, right? We still need to build these images. We may not use Docker to run them, but we need it for building it and for the whole developer experience. That is a huge, huge advantage and a huge asset to leverage.
And then the other one is Docker Hub. In the Node.js community, there’s NPM. I mean, they went and they just turned that into a business, a VC-funded business. I mean, it’s probably not the most successful business and they’ve had their ups and downs and whatnot, but they have made an entire business just out of being an artifact repository, the de facto artifact repository for Node. And Docker still has that with Docker Hub.

Jon Christensen: That’s right.

Chris Hickman: That’s worth a lot there. You go look at the companies that really want to round out their suite and the companies that really, really care about developers and the developer experience. And so again, I’ll [crosstalk 00:00:21:25].

Jon Christensen: One of them comes to mind.

Chris Hickman: Like Microsoft, come on already. Just do it. Just pull the trigger. This has got to be such an easy acquisition to make.

Jon Christensen: It’s the easiest decision in the world, Chris. Wait 11 months and then do it.

Chris Hickman: Yeah, I mean-

Jon Christensen: No, no, seriously. Why do it today when you could wait 11 months and do it for a 10th the cost and still have all the same people?

Chris Hickman: I know, but still it’s a rounding error for them. Just end the misery.

Jon Christensen: Yeah, that’s true.

Chris Hickman: We’ll see.

Jon Christensen: Yeah, we’ll see what happens.

Chris Hickman: But definitely some very big, interesting changes for Docker, the company and it’ll be interesting to see how it plays out.

Jon Christensen: Right on that. Well, thank you, Chris for joining us live today from Seattle.

Chris Hickman: Yeah. Thanks, Jon Christensen.

Jon Christensen: Talk to you next week.

Chris Hickman: All right. See you.