Cloud: From Hype to Judgement

A few years ago, all the excitement about cloud felt – looking back – a little naive. It was billed as the future of infrastructure, with promises to simplify, accelerate and liberate us from the tyranny of data centres. But just like with blockchain and Big Data, the hype was louder than the reality.

Seven years on, we’ve all learned a few things. Some of them were obvious in hindsight – but at the time, it was hard to challenge the narrative, especially when many had no frame of reference or, in some cases, expectations that weren’t rooted in reality.

Some firms went all-in and haven’t looked back. One CIO put it nicely: “We’re still spending the same, but I never have to talk about infrastructure anymore.” That’s a quiet kind of progress. But for others, it’s been less of a revolution. The big difference? Business – and by extension, IT – complexity.

The firms seeing the biggest gains from cloud tend to be the ones with relatively simple estates. That makes sense. If you can move most of your applications and data to a single provider, the benefits stack up quickly. No data centre management, no hardware lifecycle headaches, no versioning sprawl. That’s a win.

But complexity changes the equation. For larger or more regulated organisations, with deeply embedded legacy systems and interdependent workflows, cloud doesn’t remove complexity – it just relocates it.

Some things have genuinely improved. Take elastic compute: what used to take hours can now be done in minutes. That’s great – until you see the invoice. Speed has a price. If you need the result urgently, you pay for ‘express delivery’. If you can wait, opt for ‘economy’. It’s a useful flexibility, but few have mastered how to manage that trade-off at scale. Cloud makes cost variable, but variable doesn’t always mean controlled.

The same goes for storage. In the beginning, the instinct was to put everything in the cloud. But reality bites. Frequently used, high-volume data – especially in a multi-cloud setup – can be costly to keep off-prem. That’s why a hybrid model often makes more sense. Keep recent, high-traffic data where it’s cheapest and fastest to access. Let older, less-used data graduate to the cloud over time, using tiered storage to keep costs in check. It’s not rocket science – but it does require visibility, discipline and a clear data strategy.

The cloud is incredibly valuable and has become an essential part of the modern tech landscape. Where many started by considering cloud was the goal in itself: the “why”; it is actually another tool that we have at our disposal: a “how”. Like any infrastructure, the question now isn’t whether to use it, but how well you manage it.

And that’s the shift. Cloud has gone from being the headline to being the context. It’s a tool, not a miracle. If you treat it like a magic solution, you’ll be disappointed. If you treat it like an evolving platform and build your usage around clear use cases, you’ll see results.

The bottom line? Cloud still matters – but how you use it matters more. And in a world where everything is available at the click of a button, the differentiator isn’t technology. It’s judgement.