How to Build a Successful Company in an Era of Disruption

This is a new skill for companies to learn. Especially for infrastructure or application companies, you almost have to ignore the traditional selling chain. That chain, where a seller sells to a seller who then sells to a buyer, is now broken. SaaS began this disruption, but AI breaks it even further. This means product managers and others defining products must now focus on directly reaching the end user.

VMware's 20-year history saw its first decade focused on disruption and growth. The second decade involved others attempting to disrupt us. This is essentially VMware's history. Our industry, particularly on the infrastructure side, has what I call three or four 'weapons of mass disruption.' One is introducing new software that creates a new abstraction or usage model. When everything aggregates around this, you can successfully attract a new class of users who were not previously consumers. This aligns with the classic innovator's dilemma or a new business model.

You get into a mode where you're good at the business math; everyone knows the company's gross margin. However, you've lost the business's soul: are you innovating with impatient velocity to continuously leapfrog? When that happens, you must hit the reset button. About five or six years ago, that happened to us, and we had to reset. Let me tell you what needs to happen when you hit the reset button: you must ensure you have people with a founder's mentality.

The concept of being 10x better is counterintuitive for large companies. They often focus on catching up, but no one wins by playing catch-up. Instead, approaching from a different angle and being at least an order of magnitude better is crucial. You must constantly push for this and remain highly skeptical. Many claim to be 10x better when they're only 15% better, and a 15% improvement won't displace an incumbent.

We structured it like a two-pizza team initially, starting very small. This team has agency and air cover from the very top of the food chain. You must protect this team because there will be many who argue against it, insisting it should be part of the core. You have to ensure they are protected and focused on driving the initiative. They might excel at getting a version one product to market, but they won't be good at getting the entire sales force, like Cisco's 17,000 sellers, to adopt it.

One of the best pieces of advice I received when I took this job came from a board member, the ex-CEO of Northrop Grumman. He pulled me aside and said, "I'm going to give you one piece of advice: Don't delegate the storytelling to anyone in the company. You do it yourself." He explained that you need one voice. It's not about being the best presenter, but if one person doesn't tell the entire story, who truly owns it? You start to fracture the narrative, and once that happens, you lose control. The story is incredibly important; Apple is the best in the world at this.

I believe our self-perception is the critical infrastructure for the AI era. We need to consider where AI is currently constrained. It's limited by power, compute, and network capabilities. If a data packet is delayed reaching the GPU, the GPU becomes idle, which is like burning money, especially during training runs. Therefore, ensuring a highly efficient packet flow is crucial. This means low-latency, high-performance, and energy-efficient networking infrastructure is super important for AI training.

I use a six-part formula for building a great company, prioritizing what's truly important. First, and most crucial, is timing. Get the timing right now; don't fight the megatrend. Ensure everything you do leverages AI as a tailwind, or you won't succeed. Second, target a very large market that you can attack incrementally. Trying to control the entire market at once is difficult. Focus on a large Total Addressable Market (TAM), addressing it step-by-step, and ideally, create the TAM rather than pursuing an existing one. Third is the team. While some ask if the team trumps the market, I believe the market always trumps the team.