Juniper Crashes Brave New Worlds of Networking

The veteran hardware maker's CFO tells how his team is easing the company's entry into software and cloud technology.
David McCannMarch 10, 2017
Juniper Crashes Brave New Worlds of Networking

Juniper Networks has made its living for 20 years by selling networking hardware like routers and switches, mostly to telecommunications companies.

Ken Miller

Ken Miller

But while that business still provides the bulk of the company’s revenue and must be serviced and maintained, it’s no longer growing. So Juniper, like other mature technology firms in general and networking companies in particular, is compelled to transform itself.

The most prominent new directions are a hard charge into cloud technology and a less-developed but still ambitious foray into software. The company’s five-year plan foresees additional major initiatives.

How Startup CFO Grew Food Company 50% YoY

How Startup CFO Grew Food Company 50% YoY

This case study of JonnyPops’ success highlights the unusual financial and operational strategies that enabled rapid expansion into a crowded and highly competitive frozen treat market. 

It’s a whole new game for CFO Ken Miller, who came to Juniper from Ernst & Young in 1999 as a young accountant. At the time, there were only five finance and accounting employees.

In the years since, with Juniper ever the primary challenger to networking-sector leader Cisco Systems, the staff has swelled by about 100 times. For his part, Miller has climbed through an array of senior positions in the finance organization and in business units, culminating with his promotion to finance chief a year ago.

He recently spoke with CFO about disruptions in the networking industry, Juniper’s transformation, how the finance team is facilitating the shift, and his career. An edited transcript of the discussion follows.

You’ve been at Juniper for 18 years, so you know the company cold. At the same time, you don’t bring perspectives gained by working for other companies. How do you see that?

That’s spot on. I’d say I know how Juniper works as well as anybody in the company, from an operations standpoint and how to get things done. I know exactly who to talk to at whatever grade level they are to make something happen.

But, yes, my blind spot, if you will, is that I don’t have a lot of experience to say, “Well, we used to do it this way at this other company where I worked.”

I try to compensate for that by being open to new ideas. I don’t view myself as a glass-tower CFO who just tells people what to do because I know the right answer. It’s more about, “What’s the problem? How are we going to solve it?” I think some CFOs have more answers because they’ve seen things from different perspectives.

What are some things you’ve opened your mind to?

For example, some types of commercial arrangements. There’s a lot going on at Juniper about how we price and sell our products. We sell boxes — we ship them then recognize revenue. But now we’re looking at pay-as-you-go models and subscription models.

Transitioning our invoicing and revenue recognition to a more software-based model is something I haven’t done, because I haven’t worked for a company that’s recognized revenue that way.

Another area is partnering opportunities around the globe. I don’t come at those opportunities with a “here’s how we’re going to do it” approach.

What’s behind the company’s transformation initiatives?

In technology, you need to evolve or you will get taken out. The networking industry is going through a period of pretty disruptive change. There’s a lot of talk about the boxes being supplemented or replaced with software, virtualization technologies, etc. The market for boxes and hardware will continue to compress. There’s a fair amount of pricing pressure in that business already.

How do you replace the functionality of hardware with software?

Well, you don’t replace the physics. To actually move a packet, you need electronics. The software allows more customization and a new level of sophistication. If you put some virtualized functionality on the box, you can do multiple routes at the same time. You automate more rules rather than just connecting a wire from point A to point B.

What are you and your team doing to facilitate this transition?

First of all, Juniper doesn’t have a problem coming up with new ideas. Like many companies, we have a hundred pounds of potatoes but only a fifty-pound sack. So part of our job is to filter all the new ideas in R&D and business development, and even go-to-market ideas. We don’t want to build widgets that have awesome technology but little demand.

But we also have to make sure we don’t over-rotate. Our goal is to have balance between how much we’re investing to keep the business running, basically sustaining today’s products and delivering this year’s products, and how much we’re spending for the bigger bets — call them year-three or year-four technologies.

In planning, one thing I do is carve out some money that I don’t hand out. I call it my hold-back. The CEO knows about it. It’s for innovation. We want to make sure we don’t spend all our money on keeping the lights on.

What’s the process of making the allocation decisions like?

We have our Horizon 1 stack, which is keeping the lights on and R&D for projects this year and next year, and our Horizon 2 stack for the longer-term projects. You have to optimize within each stack and also among the stacks.

We try to make the process data-driven, but data changes. We might think the industry is moving in one direction, but it might pivot in a different direction. So, another part of our job is to make sure we have a “fast fail” on the big bets — to scrap them sooner if we think they’re not going to pay out. We’ve spent a fair amount of money on scrapped projects, and they will continue to happen, but it’s my job to limit that to the best of my ability.

I’m driving an internal change to not make [these decisions] a once-a-year project but to make the process more routine, do it quarterly.

Where does the company stand on the path of software development?

About one-third of our business is services and two-thirds is products. Almost all of the latter is still hardware. Our software business is 3%. I’ve said that three years from now it will be 15% of our revenue. That’s pretty rapid growth.

Is it? Three years seems like a long time in the technology field.

This industry evolution will continue for at least a decade. A lot of that is because customers aren’t ready for it. Don’t think of this as a disruption that’s going to instantly take over and be the new way of doing things. Networks are large capex projects, and companies don’t just flip them overnight. They really migrate at their own pace.

What are you doing regarding cloud technology?

The cloud is the biggest driver in our space, both the public cloud and the private and hybrid clouds that enterprises are building. Telecommunications is still our biggest vertical, and we’re focused on maintaining our revenue there, but it’s not growing. Whereas we saw 25% growth in our cloud vertical last year.

Think of a cloud as a large data center. Inside that building, a tremendous amount of high-performance networking is required.

And think about the customers, like Amazon, Google, and Facebook, that are building very large clouds. They can show we’re relevant as they continue to build out that infrastructure. And we’re also helping enterprises, large ones in particular, to build their own clouds.

So there’s no room for growth in your telecom business?

A little further out there will be. Between now and the advent of 5G, which I think will happen in 2019, we believe the traditional telecom companies — AT&T, Verizon, British Telecom — will have to turn their legacy networks into more agile ones that leverage automation via software as well as hardware. We will help them do that. It’s a big opportunity for us.

You’ve worked for three CFOs at Juniper. To what degree do the ways you think about and do your job derive from what you saw them do?

Clearly I was influenced. When I started at Juniper I had very little understanding of the CFO role. I viewed it very much through an accounting lens — close the books and move on.

The first CFO I worked for, Marcel Gani, was much more than a CFO in my mind. He really functioned as the COO. While the CEO had the vision and a lot of the customer relationships, Marcel made a tremendous number of business decisions on things like how to go to market.

I didn’t know operations, business partnering, and strategy were CFO roles. It all seemed brand new to a 29-year-old accountant coming out of Ernst & Young.

Most CFOs have an operational role today. What’s different about my role in operations is that I actually did it here at Juniper. I was actually an operator, not just a person telling people to operate better.