Hobbled by partners Dell and NetApp, where does Cisco go from here?
A Trump-repatriated cash pile could give the borg acquisition hunger pangs
Analysis While HPE and Dell are concentrating on being better on-premises data centre suppliers in a hybrid cloud world, IBM on becoming a cognitive computing software supplier, and both Oracle and Microsoft on a move towards cloud, what is Cisco’s gameplan?
Can the networking tiger from Tasman Drive evade evolutionary pressures threatening to make it as relevant as the Thylacine, the Tasmanian tiger that’s now extinct?
These questions have been hanging over Cisco ever since it started trying to get into storage with its farcical Whiptail acquisition in late 2013.
Every analyst and their sibling knew that on-premises major IT supplier dominance and growth meant having a strong presence in each of the four legs of the enterprise data centre table: servers, storage, networking and system software. Strength in adjacent markets such as security and collaboration was good but not core here.
Every one of these four legs is and has been threatened by commodification, which is leaching away value from costly proprietary products:
- Servers are just Intel-powered boxes
- Storage arrays are just collections of disks and/or SSDs driven by SW running Intel-powered controllers or in servers
- Routers and switches can have SDN (software-defined networking) running in white boxes duplicate their functions at lower cost
- On-premises system software supply empires are threatened by the public cloud providing cheaper run-time environments and substitute software
After 2010 or thereabouts, Cisco entered the server world with its UCS products and then partnered with EMC in the converged infrastructure Vblock business and also with NetApp in the FlexPod reference architecture area, twinning its servers and networking with EMC and NetApp storage. Clearly then it saw the use of having combined server, storage and networking components in a system offer to customers.
Convergence narrows the jaws
As time progressed, the power of converged component-based system offers strengthened - witness hyperconverged systems like those from Nutanix. So too did the possibility of customers moving their data centre workloads to the public clouds of Amazon and Google. As Microsoft, and then IBM and Oracle saw the strategic imperatives and built their own public clouds, then black pods of cloud death started targeting on-premises data centres. The existing vendors saw that convergence could help keep business customers on-premises as convergence equalled lower costs and simpler operations.
But they also saw that the enterprise data centre market, meaning Intel-powered boxes at its core, was declining and that their growth depended mote and more on taking share from competitors or going someplace else to do business. So we have:
- IBM getting out of hardware, apart from mainframes (near-enough monopoly) and focusing on software, services and the cloud
- Oracle building its own public cloud and drawing in its HW investments as the light from Sparc dims
- HPE getting out of PCs, printers and upper-stack software as it focuses on shipping data-centre HW, meaning servers, storage and, we think, networking
- Microsoft getting into the public and hybrid clouds with its Office 365 bridge strategy
- Dell buying EMC and getting rid of non-core SW like the content management stuff
Since getting into servers and burning its fingers with Whiptail Cisco has made no strong moves in storage or in platform software, the HyperFlex hyperconverged stuff being too lightweight to make strong progress and, being crude, being an exercise in server sales preservation.
Its networking empire has held pretty firm against depredations from SDN on the one-hand and startups like Arista on the other. We might characterise Cisco as the NetApp of a networking world in which there is no EMC. Cisco is not an EMC though, it’s not an eater of its own products. At its heart, EMC took on board the idea that only the paranoid survive whereas Cisco seems to have a streak of complacency running through it.
Cancerous on-premises data centre growth stopper
EMC went through a corporate crisis for which CEO Joe Tucci provided the rescue and after which he never let EMC relax into complacency, indeed pushing for and driving it into being acquired by Dell. Both Joe T and Michael D saw the big C, the cancerous on-premises data centre growth stopper that said, in that classic corporate cliche, get big, get niche or get out. They went for getting big. As did HPE (in data centre kit supply) while Oracle and IBM decided to become on-premises data centre growth shrinkers themselves and took up the big C habit.
Which leaves us looking at NetApp and Cisco. NetApp is aiming, it seems to us, to grow by converging on-premises and cloud storage, and partner with Cisco (servers and networking) and the cloud suppliers. It’s making noises about getting into hyper-converged systems and, if it does, then a key question will be which server vendor it partners with.
It has the existing FlexPod business and won’t want to destroy that, just as Dell doesn’t want to destroy the Cisco-powered VCE business. And yet a UCS-powered NetApp hyperconverged system will compete with Cisco’s own HyperFlex hyperconverged system. That’ll be a nice market positioning problem.
For both NetApp and Dell, partaking of Cisco networking and servers has become addictive, and they fear the withdrawal symptoms of going cold turkey.
But Dell has the wherewithal to wean itself off Cisco, with its own servers and nascent networking. NetApp, however, is more deeply hooked - Cisco being its server crack and networking cocaine.
This is a two-way street, though, and a cocaine supplier also needs its customers. For Cisco to go more deeply into on-premises data centre equipment supply, and compete with HPE and Dell, means it has to strike out with its own storage and/or become a platform software supplier as well. It has to weaken its links with Dell and NetApp.
William Blair analyst Jason Ader has taken a look at Cisco and says: “With the IT chess board changing rapidly, it is increasingly unclear to us what Cisco’s end-game is — and the proverbial clock is ticking. Does the company want to double down on data centre infrastructure a la Dell-EMC or HPE? Or does Cisco want to make a hard pivot to software, becoming less reliant on traditional hardware? Maybe its future lies in cornering and consolidating the fast- growing security market, where it already does well but currently represents a scant 5 per cent of total revenue?
Ader says: "What does not seem to be working, in our view, is Cisco’s current, piecemeal strategy...the fact remains that greater than 60 per cent of product revenue (and an even higher percentage of services revenue) is still tied to the slow-growth, secularly challenged areas of switching and routing, which leaves Cisco in a vulnerable position.”
Ader is not that impressed with Cisco’s software moves: “We estimate that less than 10 per cent of Cisco’s sales are pure software today. While the proposed AppDynamics acquisition looks interesting over the long term, it is too small to move the needle anytime soon, and synergies with the rest of Cisco’s business are not obvious.”
He sees four potential paths forward:
- Spread your bets (appears to be the current strategy)
- Double down on data centre infrastructure
- Hard pivot toward software
- Security, security, security
He is a financial analyst after all, though, and has become rather interested in US president Trump’s hints that US companies could repatriate overseas cash piles without paying huge tax penalties. Here’s what he thinks:
The good news is that should repatriation come through, Cisco would be in a position to execute on a strategic transformation that we believe is necessary to change its fate. Ultimately, to execute on this transformation, we believe the company needs to place needle-moving M&A bets with a clear strategic objective, instead of its current, incrementalist approach.
Buy, buy, buy
He says: “Cisco has used its dominant position in networking as the nucleus of its portfolio expansion strategy — and this has worked well. Yet, we would argue that the current industry shifts toward the cloud and software-defined models are much more tectonic and transformational than what we have witnessed in the past 15 years. As a result, we do not believe the old Cisco playbook of numerous tuck-in acquisitions combined with a few larger bets in emerging product areas will work.”
We believe Cisco needs to embark on a fundamental business metamorphosis. This will require needle-moving M&A bets with a clear strategic objective, in our view, instead of the piecemeal approach of the past.
Ader is getting aerated about Cisco’s acquisition potential, and identifies several potential buys.
In the data centre infrastructure area he sees NetApp, Nimble Storage, Nutanix, and Pure Storage as possibilities. Red Hat, ServiceNow and Splunk are flagged up as possible software buys. In the security area he mentions Check Point, Palo Alto Networks, Proofpoint and Symantec.
Will any of these happen?
In our view, if the EMC-engorged Dell and the better-focused HPE start getting market share against Cisco, and if the public cloud move becomes more insistent, and if networking challengers take more business away from Cisco then, yes, one or more will. ®