Where will VMWare be in 5 years?
For many years, VMWare was practically synonymous with virtualization. It provided multiple virtual servers on a single physical server, with a great feature set, good (for its time) management interface, and enterprise customer support.
Lately, VMWare has been under threat, primarily due to 2 factors:
- Public Cloud: When deploying to the public cloud, customers don’t just wash their hands of managing compute hardware, storage and network. They also wash their hands of the “hypervisor”, or virtualization layer, that is the key to VMWare’s products. Amazon certainly doesn’t use VMWare – they use open-source Xen, modified to Amazon’s requirements, with a custom-built interface – Microsoft’s Azure is built on their own Hyper-V, and I would be stunned to find Google using anything other than optimized open-source.
- Containers: Containers use native Linux capabilities to virtualize the operating system, rather than the hardware. With new support for containers from Microsoft, the need to virtualize hardware, VMWare’s specialty, becomes less and less.
Market realities and responses are one important way to predict a company’s future. A second, albeit lagging, indicator is financial performance. A company whose future is weak may have strong current income, while one whose future is extremely strong may be operating at a large and growing loss (see, “startups”, almost all of them).
Yet perhaps the most telling indicator is customer perception.
As a general rule, if customers really like a vendor, the vendor can get a lot of slack before they are thrown out. That slack, in turn, can translate into more revenue to build the next generation products or even entirely new business models, positive feedback from customers describing what they need, willingness to act as a test bed (or, depending on the circumstances, guinea pig) for new products and many years of longer life.
On the other hand, if your customer is at best neutral, and at times feels like you are squeezing them dry, then they will look for every opportunity to remove you from the picture, examining every potential replacement.
In the first decade of its existence, customers loved VMWare. The products saved customers enormous sums wasted on multiple servers where one would do. The available knowledge and support were excellent, and customers were happy.
Over time, that level of positive feeling has turns into an equal amount of negative feeling. While few hate VMWare the way they once hated Microsoft, many still hate Oracle, and everyone hates Nuance, the emotional relationship is somewhere in the range of “mild dislike” to “hatred.”
Part of this is because customers have gotten quite used to virtualization. It no longer is “VMWare Magic,” but instead a basic service requirement. When that basic requirement costs a lot of money, with prices rising on a regular basis, they quickly become a target for customer ire.
What should VMWare have done?
- Get ahead of the technology. Some people, somewhere in VMWare’s vast R&D, knew about containers and the public cloud and what they meant. Some of those, in turn, must have been knocking on the CEO’s door, yet either were ignored or were actively disciplined for troubling the leadership with trivial ideas.
- Make the API free. VMWare, in its search for revenue, charges partners a fee to access the API of its core products. This has the perverse effect of driving ISVs away.The reason many companies nowadays release their products open-source is to enable third parties to build new services on top of the product, thus creating new revenue for partners and increasing customer dependency on the platform. Charging for the API is penny-wise pound-foolish.
- Keep prices as reasonable as possible to drive up volume. They aren’t selling hammers but rather ecosystems. The more that use it, the better the long-term revenue.
Will VMWare survive the coming armageddon? I truly do not know. They do, however, need a lot more than another feature or major product version that is more of the same.