For years, cloud computing has been hailed as the revolutionary future of technology—merging innovation with the promises of cost savings and agility.

Today, the wave of enthusiasm for public cloud — those powered by AWS, GCP, Azure, and others — is dissipating.

Many enterprises are questioning whether public cloud solutions are as effective, cost-efficient, and strategically aligned as advertised. The anticipated productivity gains haven’t always materialized. Concerns about vendor lock-in are rising as companies become wary of being beholden to a single provider’s ecosystem.

Cloud-first and cloud-centric business strategies are now being reassessed.

In turn, on-premises hardware data centers and colocation solutions are making a comeback.

A high-profile example is 37signals, the makers of Basecamp, Hey, and more. They “de-clouded” their hosting early this year, leaving AWS and GCP behind.

Their result?
By managing their own servers, they are already generating significant benefits for their business, especially cost savings.

How will the cloud providers respond?

It will start by reducing costs and embracing openness. We will see more promotion of interoperability and how they are helping their customers avoid ecosystem lock-in.

And, of course, there is AI.

Integrating AI into cloud platforms is expected to drive growth, but this may only be a reprieve rather than a long-term solution. As Apple, Microsoft, and Google are showing their ability to run AI on laptops, the ability to run AI through on-premises hardware is achievable today.

When the hype fades, cost-effectiveness and practicality are always the keys to long-term success (right, Airbnb?).