It’s cheaper for most businesses to rent a cloud than own one. But it’s even better to be one.
That’s why Apple, Microsoft, Amazon, Google, and Facebook are the five most valuable companies in America. Each is worth over $1 trillion. Each put the “ante” of $1 billion/quarter into cloud construction in the last decade. Each continues to pour $3-5 billion more of cash flow into their networks every 90 days.
Policymakers focus on the services these companies provide, but it’s the cloud that makes them powerful. Microsoft isn’t software. It’s cloud and software, which lets it take cash from competitors. Google isn’t services, it’s services and software, which also let it take cash from its competitors. Only Facebook, so far, is refusing to rent capacity, and even it sells access to what’s becoming a global fiber network.
That’s where we are today. Where will we be in 2025, if man is still alive?
Big software companies, including IBM, Salesforce, Oracle, and SAP, have finally gotten on the cloud construction train. They don’t control all their processing, but they’re off-loading some of their capacity into their own data centers.
This is just the start of a long trend that will change the cloud landscape. To big companies the lesson is clear. Become a cloud or remain subservient to the Cloud Czars.
The Cloud Czars have sought to forestall this with “industry clouds,” loaded with software that appeals to specific types of businesses. Google has sought to forestall it by buying cloud market share. This has its stock rocketing even though (alone among the Czars) it’s losing money in cloud rental.
But as numbers get bigger, more companies are going to want to move away from the Czars. This will start with “hybrid clouds,” cloud-based data centers, built to cloud standards, that connect to larger clouds. Building these things could prove a great business for companies like Dell. They can be anywhere, thanks to fiber cables. Facebook dominates in the developing world despite having 10 of its 15 data centers in the U.S.
The next step will be a balancing act, as scaled companies seek to bring more of their cloud-based services in-house. This is only possible once you have the cash flow to sustain the investment, and once all your software is cloudy. For most this will remain a pipe dream. But by 2025 most large companies will have their own data centers, built to cloud requirements. They will be doing internal cost comparisons on every big project, gradually bringing more operations in-house.
What the Czars will do in response is to go “up the stack,” offering higher value services. Amazon started with pure infrastructure, called “bare metal” in the trade. Within a few years, either directly or through an acquisition they’ll be offering office and accounting suites, more like Alibaba.
What does this mean for policy makers? It means they’re looking at the wrong thing. They’re focusing on services, on breaking up companies like Facebook based on those services. Doing so risks the cash flow that makes low-cost cloud possible. Instead, they should foster multi-cloud and hybrid-cloud growth. This will bring things into balance organically, between the Czars and those they serve, or between services and the Czars they serve.
