Why hyperscalers are a menace

Hyperscalers are cloud service provider such as Google, Alibaba, Amazon Web Services or Microsoft Azure. They provide infrastructure, computing power and a variety of services. Companies, states and organisations can use their services and are billed by usage.

The name hyperscaler implies that the more organisations are using their services the cheaper they become. This effect causes organisations to migrate their infrastructure and services to their hyperscaler of choice.

At first thought it seems like good practice, however, there a few issues that make these hyperscalers a menace.

Monopolies

It is well known that a free market with monopolistic actors causes a welfare loss. Every hyperscaler currently is aiming for a global monopoly position and everybody seems to turn a blind eye. Companies that try to create a monopoly in a free market must be regulated.

Interoperability

If a organisation has its infrastructure run by hyperscaler there is currently no way to switch to another provider, although the required standards to make interoperability possible do exist. This vendor-locking is severe and limits the possibilities of choice (aka freedom).

Vertical Integration

This is the most important argument. When comparing the stack of value (networking, computing, monitoring, containerization, clustering, security, …) that is used to provide cloud-computing, it is all provided in vertical manner. A provider controls the whole stack of value. The opposite is a horizontal integration would be much better.

Bad: Vertically integrated hyperscalers

Good: Horizontal integration

In a horizontal integration different provider compete for each value layer. Guarantees and contracts between the value layers are managed with service level agreements (SLAs). Moreover, if competition is given interoperability between providers becomes a must.