Steve Mullaney and I started our careers in the communications industry at the same time — 1992 — and in the same place: Silicon Valley.
Back then, the World Wide Web wasn’t a thing (it didn’t arrive until 1994) and the buzz was all about 10Base-T Ethernet LAN hubs, which delivered a mighty 10 Mbps of shared bandwidth over copper cable.
Steve oversaw the design of Ethernet hubs for Synoptics; my job was writing about them for a musty old print magazine (remember them?) called “Data Communications.”
It’s easy to be dismissive of 10Base-T today, but back then it was revolutionary stuff, and Steve’s first job marks the beginning of a serial trend in his career — a preternatural ability to identify emerging communications transitions and pick a winner to work for within that space.
Doing that once is considered a good trick in an industry where 19 out of 20 startups fail. To do it as consistently as Steve — at Nicira, Force10 and Growth Networks, to name three — is genuinely remarkable.
Given his history, it's not surprising to find Steve today overseeing Aviatrix, a private company smack in the middle of multi-cloud — the most exciting and chaotic segment of the cloud infrastructure market.
I caught up with Steve on a Zoomer last week. “Ask me anything,” he said. “I'm very transparent.” So I did.
Steve Saunders: Wow, what a career you're having.
Steve Mullaney: I hang around just long enough, I guess.
Saunders: You retired for five years. Why return now? And why at Aviatrix?
Mullaney: I was on the board of Aviatrix and at the time they were barely surviving. You gotta remember five years ago, there was no multi-cloud. There were no enterprises moving into cloud. There was just DevOps [teams] going around spinning up workloads in AWS [Amazon Web Services]. And they asked me to become CEO and at first, I said ‘absolutely not. I’m retired.’
But right about that time the Cambrian explosion happened: cloud’s version of going from single cell organisms to multi-cell complexity.
Saunders: Cambrian explosion?
Mullaney: Enterprises have always moved like a herd. They did it with the move to IP and with the move to client server architectures away from mainframes. How did IBM, DEC and all those guys lose their business? Because transformations go from zero to one on a Tuesday morning at nine o'clock. And so, I'm on all these boards — all infrastructure companies — and it was clear to me that the conversation had just completely changed.
You don’t know what the next computing model is going to be until it arrives. Nine years ago, if you’d told an enterprise they were moving to the cloud they would have laughed. But four and a half years ago, every single one of them said, ‘No more wild west, no more shadow IT spinning up workloads in the cloud. I need visibility and control.’ And I just thought, this is too good to pass up.
Saunders: How big did you think the opportunity was?
Mullaney: I knew this could be the biggest thing I've ever done by orders of magnitude. It was clear someone was going to become the equivalent of Cisco in the last great computing model. And I knew it wasn’t going to be an existing player, like Juniper, because it never is in that kind of a transformation. So, I came back for three words: ‘Why not us?’ At that time, we were a few million in ARR [annual recurring revenue], we had about 100 customers that you've never heard of. We were filling holes. But we were cloud native. We were in there.
Aviatrix and IPO prospects
Saunders: The goal for Aviatrix is obviously an IPO. But the last three years have seen a lot of headwinds preventing new public market entries. What’s your view on Aviatrix’s IPO outlook?
Mullaney: We changed our approach around July . Before that the world was different; the priority was to grow at any cost. Our last round we raised $200 million at a $2 billion-valuation (we still have $140 million in the bank, by the way).
So, back then, [it was] who cared what you were spending or what you were burning? If I needed more money, you would just go to the ATM machine, and you’d pull more out. I could raise $200 million more at $4 billion, right? Take some minor dilution, put another $200 [million] in the bank... Just as long as we’d just grow, grow, grow, grow, grow.
And then overnight, you [could] see that that wasn’t going to work.
Now, you can't go public if all your costs are going up. What you need is moderate growth, and profitability. We’re at $75 million ARR right now, this month. So, we know we're going to be cashflow positive in 18 months. And the reason is we have an 85% gross margin.
Saunders: How have you managed that?
Mullaney: Because we’re not a SaaS [software-as-a-service] company. You buy our software and run it on your instances in the cloud. It's the customer's infrastructure.
Saunders: What’s the main benefit of that approach to your customers?
Mullaney: Visibility. And control. Cloud was originally designed for DevOps people and SMBs [small-medium businesses] who don't care about the infrastructure.
Enterprises are a whole different ballgame. They need the agility of the cloud for the developers, but they also need visibility for troubleshooting, and to control the security and performance and traffic engineering and application performance and costs, and you name it.
Networking is the most strategic piece of infrastructure an enterprise has, and enterprises don't like the cloud providers because they get no visibility and control. When there's a problem, AWS says, ‘Well, I don't think there's a problem, Steve,’ and you can’t prove it because they don’t give you any visibility.
And that’s fine if you’re just cobbling some scripts together to run a couple of applications over AWS or Azure.
But as soon as you’re running mission-critical, enterprise-grade, business-grade applications, and across not one cloud, but multiple clouds, and downtime is a million dollars a minute, now you’re something legitimate. That's where we come in.
Saunders: One of your biggest competitive advantages is your people, and the fact that you’ve managed to attract so many top-tier developers. How do you do that?
Mullaney: Well, Google is imploding right now, which is fantastic. Our head of engineering came from Google. We were getting a bunch of great Google people, even before their recent layoff. Just about anybody in networking at Google has come here. We have 375 people now, 10 times what we had when I joined, and the revenue is up 25 times.
Saunders: It always comes down to the people, doesn’t it?
Mullaney: It does, and there’s another people trend that I think is super interesting taking place within the enterprise. Four years ago, we typically sold our services to cloud experts. Usually, they were part of a small scouting team inside an enterprise customer, who had been given the job of figuring this cloud thing out.
But since then, those people have been through the process of trial, error, test and refine, and they’ve signed off on the move to cloud. And they’ve done their job, so now they want to transfer cloud networking to the on-prem networking team, and they look around and they find someone who has networking in their title, and they say, ‘Guess what? You now own cloud networking.’ And often those people can’t spell ‘cloud.’
Those are exactly the people that Aviatrix helps. We go in and we say: ‘Hey, it’s O.K., we know cloud, we're bilingual, and multilingual, we speak all the clouds.’ And we put a warm blanket around them, give them training, and we make it easy to move the software to one common architecture, to enable these advanced networking and security and telemetry services, and to do all the things required to turn them into cloud heroes at their company.
We want to elevate these heroes and promote them. Nike has its athletes. We have our cloud network architects.
Saunders: This is a crazy industry that we work in. It's nothing like the industry that you and I started out in. It's completely chaotic, undefined. Do you enjoy the chaos?
Mullaney: I love creating things and creation is not orderly. One-to-n is orderly; that's when you create order out of chaos. Zero-to-one, creating something out of nothing, is chaotic. I like that. When you're in the zero-to-one stage, you don't create a yearly plan and then go away and execute. Because next week, you learn something new, and it changes.
When Gartner comes out with a Multi-cloud Magic Quadrant early next year, that's when we'll check the box and say, ‘Now we’re officially in the one-to-n stage for multi-cloud networking.’ We’re not at that point yet.
This interview was edited for clarity and brevity.
Have we mentioned that we love mail? Send us a letter here.