Intro
Now that my wardrobe is starting to look less ‘corporate drone’ and more ‘traffic cone enthusiast’ thanks to my recent move to Pure Storage, I’ve found myself forming a more holistic opinion about my former employer, Broadcom. If you’re expecting a fiery exposé from a disgruntled insider, I’m going to have to disappoint you. Think less ‘scorched earth’ and more ‘thoughtful observation.’ Most of what I’ll touch on has been addressed in the media if you follow the industry, but I do think people are missing the more historical perspectives on this industry shift. What I am here to do is offer a slightly less dramatic, but hopefully more insightful, look at Broadcom’s moves. Consider it a 16 year historical perspective, boot-and-suspenders trenches of information technology.
The First Two Competitive Waves
While there are undoubtedly competitive pressures I wasn’t privy to, my time in the trenches of IT has shown me we’re currently riding the third wave of change for VMware. I started at Dell in 2009, but it was around 2011, leading up to the Windows Server 2012 announcement, that Microsoft made a concerted effort to challenge VMware’s dominance. At the time, VMware was perceived as vulnerable. Under the leadership of Paul Maritz, vSphere licensing was based on per-socket costs and a vRAM entitlement, which led to significant cost increases for customers. Coupled with a perceived lack of significant new features, this created an opening for Microsoft. Windows Server 2012 was practically a Hyper-V love letter, heavily focused on its capabilities and positioned as a direct competitor to vSphere. They touted features like ‘Shared Nothing Migration’ and emphasized spec-sheet supremacy, with Hyper-V supporting 4TB of vRAM and 320 logical cores, significantly exceeding vSphere 5.1’s 2TB and 160 cores.
Did these maximums actually make a practical difference? No…
But did it allow Microsoft to chew into VMware’s market share? Also… no.
There are countless theories about why Microsoft’s Hyper-V push didn’t dethrone VMware. In my view, then and now, it’s the accumulation of seemingly minor details that makes the difference. Hyper-V’s central management, SCVMM, was an added expense and less robust than vCenter. The lack of enterprise-grade disaster recovery orchestration, a feature VMware offered through Site Recovery Manager, was another critical gap. And, of course, Pat Gelsinger, in what must have been his ‘easiest first-day-on-the-job task,’ promptly removed the vRAM entitlement, essentially becoming the vRAM Robin Hood, redistributing licensing costs back to the beleaguered IT practitioners.
Was vCenter that much better than SCVMM? No. But did the lack of enterprise DR Orchestration become a non-starter for IT professionals? Also no.
Organizations could have made the leap to Microsoft. Many of my colleagues, at the time, considered it a foregone conclusion. Microsoft was dangling substantial incentives in front of sales reps to drive Hyper-V adoption and displace vSphere environments. Yet, the predicted mass migration never materialized. While application development sprints forward, infrastructure change is a marathon of deliberate, cautious steps.
The second wave of competitive pressure arrived in the form of the public cloud, a force far more potent than Microsoft’s challenge. The concept of ‘utility computing’ had been circulating, but AWS transformed it into the tangible reality of ‘public cloud computing,’ presenting a compelling vision: a paradigm shift promising new capabilities and concepts. ‘Infrastructure is a headache,’ they argued, ‘slow, costly, and inefficient. Let us handle it, with metered simplicity. And while we’re at it, we’ll revolutionize your applications.’ While some of those promises, like the serverless hype, have cooled considerably (remember when that was the future?), the cloud’s impact was undeniable. VMware, along with the entire on-prem ecosystem, felt the shift. Market share undeniably eroded as customers migrated to AWS and other cloud providers. Yet, despite this, VMware still crossed the $10 billion revenue threshold. Some customers found strategic value in maintaining on-prem control, others discovered cost savings were application-dependent, and some simply required performance or capabilities the cloud couldn’t deliver. Notably, some customers, like 37signals (now Basecamp), even repatriated workloads back on-prem, realizing significant total cost of ownership (TCO) savings. Ultimately, even after significant cloud adoption, the best estimate I’ve encountered is that roughly 71% of on-prem virtual machines remained on ESXi.
VMware’s 3rd Competitive Pressure… Capitalism
So now we are on the 3rd wave of pressure to VMware and its market share. Except it’s not a competitor or a revolutionary technology, it’s an acquisition. The Broadcom acquisition has dominated IT conversations for nearly two years. Storage vendors, server vendors, backup software providers, automation tool developers – everyone is reacting to this market disruption. It’s challenging to convey to those outside IT why a piece of software, representing less than 10% of total IT infrastructure costs, is causing such a stir. Imagine your home’s entire electrical and gas system suddenly being managed by a new company, one that could change the rules and prices on a whim. Naturally, you’d consider switching. While inconvenient and potentially costly, changing your utility provider wouldn’t require rewiring every household appliance or impact your heater’s performance, prompting your family members to ask what happened. The word “commodity” get’s thrown around a lot in IT and few things actually meet the true definition of commodity at the scale of what people actually mean. There are other hypervisors, but they aren’t commodities like electricity or natural gas. And depending on your IT maturity, the level of integration with everything around the hypervisor makes the research, configuration, and migration so operationally expensive, everyone gives up at the first planning session.
While prices have surged for most customers, some have seen cost reductions. Ironically, for all the griping customers did about the ‘old’ VMware, many would now eagerly welcome it back. I witnessed VMware pursue deals they’d have been better off abandoning. While this is a generalization, not every customer’s experience, I felt VMware genuinely valued customer relationships, even when those relationships were challenging or impacted profitability. The ‘new’ VMware, however, feels decidedly more corporate, detached, and ‘take it or leave it.’ The customer pain points can be broadly categorized into three distinct areas.
- Existing VCF and high-level private cloud adoption customers on subscription licensing: These customers are experiencing roughly the same costs, or even a decrease. Their sentiment is, ‘What’s all the fuss about? This is a solid deal.’
- vSphere customers with other critical SDDC components like NSX, vSAN, or Aria Suite on perpetual licensing: These customers are facing increases ranging from 20-50%. Their reaction is, ‘This is mildly concerning, and I need to explore alternatives.’
- vSphere/vCenter-only customers on perpetual licensing: These customers are enduring cost increases that feel like a personal betrayal, ranging from 50-350%. Their reaction? Let’s just say, “Hock Tan” is now a banned phrase in their vocabulary, replaced by a series of creatively constructed expletives.
While I could delve into further customer segments, you get the picture. Hock Tan is undeniably a brilliant business strategist with a strategy that diverges from the typical Silicon Valley startup approach. I’ll refrain from offering extensive personal opinions on the CEO here, but I’m always open to discussing it over a beer.
Matt’s Takeaways: What This Means for Customers
Whether you’re a VMware customer or a partner trying to navigate this new landscape, here are some key takeaways:
- VMware is the New Oracle: Prepare for a Negotiation:
- Forget the friendly, collaborative engagements of the past. Renewals will now be a battle. You’ll need to prepare for intense negotiations, know your limits, and be ready to walk away if necessary.
- Some customers will have no choice but to migrate off the platform. Define your “deal breakers.” How will you prepare your stakeholders for these challenging renewals? These are critical questions to address.
- Foster Creativity and Exploration:
- Organizations must empower their IT teams to explore alternatives. Invest in lab environments, provide MBOs for testing, and grant them the space to think strategically.
- While some leaders may view this as a waste of resources, it’s an investment that pays dividends when your team has a realistic understanding of available options and a well-defined migration strategy.
- Start Planning Early – Really Early:
- If your renewal is weeks or months away, you’re likely already behind. Planning should begin 12 months, or even years, in advance.
- The complexity of migration depends on your IT maturity. Simple shops may find it easier, but for most, changing hypervisors is a cross-functional endeavor. Security, backup, networking, and various business units will be impacted.
- Consider the ripple effect: logging dashboards, application health monitoring, backup and recovery processes, network port discovery, and security integration like micro-segmentation. The ripple effect is by far the most important part of this whole ordeal. I’m only mentioning a fraction of the things you need to consider.
What happens IMHO
So, what’s my prediction? If you’ve made it this far, you’ve likely guessed it. I believe VMware will lose approximately 10% of its ESXi virtual machines to alternative hypervisors or the cloud. However, this loss will be offset by Broadcom’s ability to achieve operating margins exceeding 70%, creating an incredibly profitable software business that facilitates debt reduction and future acquisitions.
A telling moment occurred during Broadcom’s FY 24 Q3 earnings call, when analyst Harsh Kumar posed a critical question. You can review the transcript ( https://www.fool.com/earnings/call-transcripts/2024/09/05/broadcom-avgo-q3-2024-earnings-call-transcript/ ) to understand the investor’s perspective. The disappearance of VMware’s SaaS offerings is a direct result of this strategic shift.
While counterarguments exist, the most significant risk to VMware’s business lies in the successful execution of VCF 9.0. This release marks the first major software update after a significant headcount reduction and the consolidation of engineering teams. The scope of change is substantial, and any significant issues could impact customer confidence. However, the changes primarily focus on streamlining the platform, removing underutilized software components and services in favor of a simpler, integrated experience. If Broadcom executes effectively, they’ll minimize potential disruptions and provide customers who remain a reasonable path to adapt to the new cost structure and corporate culture.