We are in the middle of earnings season, so there is a lot of industry news to sort through. In addition to two networking vendors beating estimates yesterday (Juniper and VMWare each reported earnings above analyst estimates), the big news was that Juniper CEO Kevin Johnson is leaving his CEO position to join the ranks of the retired.
There will undoubtedly be a lot of stories about Kevin's announcement. Analysts and pundits will weigh in with their own take on what happened, why it happened, and what it means. People will point to the stock price or the most recent earnings, the product pipeline, or maybe the state of the workforce. The temptation with all of this is to play armchair quarterback and talk about what decisions should or should not have been made. It is all easy when you are on the outside. Every decision is straightforward. Every move is obvious.
But Juniper's story is actually not a technology story. And the stock price cannot be explained by simply looking at products, features, and market share.
The reality for every company in high tech (networking or otherwise) is that what we sell is not a product. We sell innovation. You see the technology is forever changing. This week's hot feature is next week's table stakes technology. At the best of times, most vendors just leapfrog each other on performance or capability. Differentiation is difficult to achieve, and even when you find it, the advantage is fleeting. When you get past the marketing, with rare exception, most vendors are on more or less the same path with more or less the same vision.
Intuitively, I think we all kind of recognize this. When the market turns on talk of a protocol (OpenFlow, VXLAN, whatever), it really highlights that we are dealing with minutia. This is why new ideas and new approaches stand out. The fact that they are different is remarkable.
So what does this all mean for Juniper?
Juniper's success is not tied to a product or a feature. It is tied to its ability to come up with new ideas and drive those ideas from concept to reality. Not once but a thousand times. The real measure of Kevin Johnson's effectiveness is not PTX or QFabric or MobileNext. Those are single instantiations of a technology culture. Kevin's legacy will be whether the workforce he leaves behind is capable of continuously coming up with new ideas after his departure and driving those to market.
The role of a CEO is not to decide what products to build but rather to pick a direction and ensure that a culture that encourages new ideas and fosters an innovator's spirit exists. On the direction front, I think the future Kevin Johnson painted is clear: software is where the action is. In that regard, he was arguably more bullish and more vocal about software in the networking space than many of his peers. He has been talking software as the future since 2009. There are nuances in software that were maybe overlooked (PC and network operating systems are vastly different), but he should get credit for getting the macro thesis correct that the future is software defined.
Where I think things are less certain is whether or not he created an inventive environment that promotes new ideas from all corners of the company. Because Juniper has such a strong familial culture, many of the most senior architect positions are long-time employees responsible for many of Juniper's products from the M40 to today. These people have invented some of the industry's all-time most successful products, so their voices are weighty. With Hall of Fame minds still around, the question is whether newer ideas tend to surface or stay quiet. One thing is for sure, it won't be networking as we know it going forward, so there will need to be an influx of new ideas, likely from people who have not been behind the networking revolutions in the past. I will say that, having been at Juniper a long time, the place is chock full of talent. The ideas are there – they just need to be given room to flourish.
We all would do well to understand that it is repeatable innovation and not one-time product creation that is our lifeblood. A new way of doing things requires new ideas, and oftentimes that inspiration comes from outside the established minds. Companies that have a solid mix of new and old minds should be able to bring new thoughts without completely abandoning a rich history of real-world experience. Without turning this into a Plexxi commercial, I will say that is one of the reasons I am bullish here.
What is next for Juniper?
This is tougher to predict, and having spent 12 years at Juniper myself, I don't know that I can be unbiased in anything I put forth. I will say that core transport is increasingly a cost reduction game because connectivity and bandwidth are not the revenue drivers of the future for carriers. With new players in that space like Compass and new products from existing players, differentiation will be tough, and that will put more pressure on pricing and margins. On the positive side, Juniper is aggressive in pursuing the disaggregation of routing functions and embracing software. Contrail looks solid, and Juniper's SDN and NFV ambitions in the carrier space could yield new software revenues to offset hardware slides.
It will be interesting to see how longtime Juniper strongholds in web-scale accounts hold up. Google's move toward white box solutions has been well documented, and the Open Compute Project from Facebook is another one to watch. These, combined with players like Cumulus and Pica8, could create additional headwinds in those choice accounts that have the software expertise to roll their own in a Linux/DevOps type environment.
On the enterprise side, things look even tougher. Arista is likely preparing for their own exit. The challenge they will have is that their margins are as high as they will ever be and growth has slowed from a few years ago. If they want to have a big exit, they could compete aggressively on price, which would drop their margins and spike growth. For investors, this would be attractive because it would mean there is margin upside (there is minimal upside now) and there is growth. Those two things would make a public offering more palatable for would-be investors who are betting on business fundamentals more than technology. But the result for Juniper would be an even more price-sensitive market in the area where it should be easiest to grow share. That would put more margin pressure on Juniper, which could be tough for a business focused on EPS (the reason behind Juniper's previous announcement of $150M in OpEx reductions this year).
Juniper appears to be prepared to take this on with its orchestration and automation capabilities. Junos remains a pioneer in the automation space, and recent announcements of Puppet integration should likely be followed by other DevOps-type integrations. This should be attractive to enterprise customers, but will it offset a bolder, more aggressive Arista? Hard to say.
So who is next at the Juniper helm?
I have no idea. You do have to wonder what Stefan Dyckerhoff is thinking. It seems strange that Juniper would lose the up-and-comer in 2012 and then say goodbye to its CEO less than a year later. That kind of change in Section 16 personnel might spook investors some. So does Bob Muglia take the reins after being the most visible Juniper executive of late? Does Stefan make a triumphant return? Do they look outside to someone else? I certainly don't have a crystal ball, but whatever move Juniper makes next will reverberate through the entire industry.
In closing, I will say that while we identify as companies, we are all individuals. I have strong relationships at Juniper still, and I do hope this transition goes well for them. While it might not be en vogue to root for your competitors, I do hope the brilliant minds I worked alongside for so many years continue to shine. Reducing everything to corporate allegiance and wishing anything else would just be inhuman.