There are many different angles to Juniper’s intent to acquire Apstra, the one this article focuses on is seeing the whole problem.
Image Source: Juniper Networks / Apstra
Introduction
As noted in previous articles on the three-olive martini model, the challenge data center/network managers face is not just the complexity in the network, but the complexity across the network, operations, services, and outcomes, framed by constraints such as budget, economics, physics, regulation, company mission, etc. A simplification to the network could mean more work/complexity for operations.
The ability to see complexity across the network, operations/AIOps, and intent/outcomes is central to the hyperscaler disruption and one of the most prominent characteristics of IT today.
Apstra is not the only company in this space, and Juniper is not the only networking company looking to play in this space, but it is an interesting new area for networking companies to play, especially as they look to solve networking/IT (1) problems for Enterprises that do not have the operational capacity and capabilities of hyperscalers, but are looking to improve efficiency and agility (productivity).
Strategy Aligned Acquisitions
It does feel like Juniper is shifting to a new phase of strategy-aligned acquisitions. Recall that Juniper was previously criticized in the 2013/2014 timeframe by activist investors like Elliot Management for having not realized value from its acquisitions. The involvement of Elliot Management in Juniper stock resulted in marginal value for shareholders, IMO. While it may be possible to look at a company from the outside and make recommendations about what is not working, it takes innovators to develop value propositions that will work, and grow the company, if you want something more than flipping a share position.
Juniper’s campus-to-cloud Enterprise AI strategy seems apparent in the Apstra acquisition. The (automated) root cause identification and analysis are well-aligned with what Mist is demonstrating in campus/branch. It does not take a rocket scientist to imagine some synergies/integration over time.
Apstra also has a full life cycle management angle as well, which goes further to the focus of this article, about seeing the whole problem, and the complexities/tradeoffs/constraints across networks, operations, services, and outcomes.
Of specific interest will be how Apstra helps Juniper move the ball forward on RIFT (routing in fat trees), what the optimal balance of autonomy is across the network and operations, what the combination of the two companies will mean for the future/growth of SONiC, and how much of the full life-cycle management the two can nail and develop the brand around.
Challenges
Integration and execution are not trivial activities. They require strategy-aligned objectives/outcomes, resources, skills, and perseverance. This is a given in any acquisition. Apstra is a Menlo-park based company, which should help on the culture front.
Multivendor solutions are tricky. As a stand-alone solution company, multivendor value propositions are natural and pragmatic. As part of a company developing its own networking products, the multivendor aspect can become complicated.
I know of at least one network equipment company who went down the road of building a “software” business, but the product managers in the software businesses weighted their investments towards the competition, when the acquiring company’s own products were not already market leaders, further hindering the growth and success of that acquiring company’s own products. Juniper, already a networking leader, though not dominant in the data center space, has less of that problem, but it is something that will need to be thoughtfully managed.
Juniper and Apstra were already working together before the intent of acquisition was announced, so that helps. My advice to Juniper is, be unashamed about creating a first-class/best-of-breed solution based on Apstra and Juniper products, while also maintaining the multivendor capabilities. Multivendor does add value to customers, to the extent it can be fully realized. The high-order bit for Juniper though, is inserting more pervasively in Data Center LAN, and opening up that growth path for the company. Ultimately, it is a question of which approach generates the most value for customers.
For companies looking to grow an Enterprise business, GTM (go to market), is a critical issue. Wall Street analysts have noted Juniper’s hiring of Enterprise sales people, but the issue goes beyond that. Each additional sales person can only add so much revenue, per head. I’ve been involved in modeling these scenarios in previous strategy work.
When I think of a company that became a channel superstar, Palo Alto Networks (PANW) comes to mind. Their evangelizing of application/next generation firewalls put them in a category of their own, that led to breakout success, despite the excellent channel operations that Cisco had and has. To get beyond Fortune 100 / Global 2000 success, Juniper needs something akin to this. For now, Juniper has parts of their Enterprise portfolio, for example Mist, that have strong growth rates.
That Apstra will not be immediately accretive was probably not great news from the perspective of my friends crunching financial valuation models, but Juniper’s stock price is currently up, despite the market as a whole being down. The bottom line for investors is that this should provide Juniper with an enhanced ability to go after Data Center LAN/switching business.
Conclusion
Market insertions don’t just happen because you roll out of bed one day and decide you would like to play in a new market. There has to be insertion opportunity, strategy, and execution. If incumbents are doing the job OK, why should a customer change? The emerging industry focus on augmented and autonomous networking, provides an insertion opportunity for Juniper in Enterprise Networking. The acquisitions appear to be strategy aligned to the bigger picture of operations/outcome complexities. So now, the execution...
Notes:
(1) The extent to which Juniper may actively focus Apstra’s technology beyond networking is not clear at this time. Likely Apstra has given some thought to composable architectures beyond networking.