Skip to main content

There and maybe back again: A cloud journey - moving out of public cloud

Has your organisation considered moving away from the public cloud?


Woman at laptop 2
Matt Larder 180 x 180

Matt Larder

Head of Cloud, Softcat

Has your organisation considered moving away from the public cloud?

Moving out of public cloud

In Flexera’s State of the Cloud 2022 report, an estimated 63% of respondents reported heavy public cloud usage. Similarly, a 2022 Gartner press release reported that “worldwide end-user spending on public cloud services is forecast to grow 20.7% to total $591.8 billion in 2023, up from $490.3 billion in 2022. This is higher than the 18.8% growth forecast for 2022.”

Are all public cloud journeys one-way?

Many organisations contemplate a move back to a traditional datacentre and/or on-premises environment – sometimes clumsily referred to as IT workload repatriation or as a colleague recently referred to as de-clouding. But is this a talking point or are organisations executing on such thoughts? In this blog post, we’ll take a look at the common risks driving this behaviour, the ease of doing this and provide some guidance for your consideration, when contemplating a move back from public cloud.

Let’s talk risk

Conversations I’ve been part of in recent years, have centred around two specific factors leading the organisation to question their sole or majority use of public cloud. These factors are failures and cost.

  • 1. Failures

Public cloud is not bullet-proof. Each of the big three providers (AWS, Microsoft Azure and Google Cloud) have experienced a range of publicised failures in the past two decades. Just like when on-premises, such failures lead the user to question the reliability of the chosen IT platform, both in the sense of ensuring they have the right controls in place within the platform, but also whether the platform itself is fit for purpose. In my experience, the frequency of failure and criticality are key reasons to make a change, and the former is typically a lot less in cloud vs on-premises.

  • 2. Cost

For as long as I can remember, organisations consistently, and rightly so, question the cost of cloud. Questions often come in the form of:

- Is another cloud cheaper – in my experience, outside of custom commercial agreements, when comparing apples vs apples, costs between clouds are not that different or rather not substantial enough to drive a change. 

The cost of cloud is much higher than expected – this is often caused by costs not being controlled/governed in an on-demand platform but on balance. Simply put, cloud is not always cheaper (or right for all workloads).

In my view, both of these are much more in the control of the user. However, there is a third question, that I’ve heard for many years but have only recently seen the impact of, which is not in the control of user:

What if the cloud provider increases costs – in most cases there is a trend of the provider maintaining or reducing costs which has always mitigated this question. But recently, some providers have been increasing costs on certain agreements based on inflationary pressures and macroeconomic conditions. Similar to the failures point, this has become a very valid reason to ensure you’re consuming the right cloud (agreement or overall platform) or evaluate alternatives.

Lastly, while these aren’t additional risks, some individuals simply get ‘itchy feet’ when they’re in one place for too long, or are inquisitive, meaning they are constantly on the lookout for change.  Similarly, some organisations have a general nervousness around putting all their IT in one place. In my experience, lots of these conversations come in waves, perhaps driven by competing offerings with the promise of addressing all your concerns, or simply by the fact that cloud is now the dominant platform, meaning it gets greater spotlight when it’s use doesn’t match up to expectations.

How easy is it and how do you go back?

I’ll start my synopsis with a question for you - how easy was it for you to get to public cloud? I would wager, at this point, there are a few schools of thought amongst readers:

- Digital natives – you probably built your businesses natively on cloud and don’t have the scars of moving IT between datacentres or to the cloud.

- Those who built new cloud native workloads or moved small footprints – understanding how to use cloud was I’m sure a challenge, as was in relative terms, migrating or modernising small pockets of IT to cloud.

- Those who consolidated entire datacentres – this group I would wager are the ones really feeling the pressure of the question of moving out of cloud.

For those of you, especially the last group, that have ever moved entire datacentres to new datacentres or cloud, this was not an easy process. It most likely took months of planning, execution, lessons learnt, all adding up to at least a one year programme, if not considerably more. And no matter the destination, the process was typically the same; understanding what you have, mapping what you wanted to move, identifying the people and tools needed to move, building programme plans, migration plans, end user cutover plans, planning for rollbacks, completing test migrations, executing real migrations in the middle of the night or over the weekend, testing your migration and fine-tuning the destination to keep your users happy.

So simply put, moving out of cloud will be as easy/hard, costly, stressful, short/long as it was for you to get to cloud initially. Is this something your business can afford to endure again? Quite possibly, if some of the market pressures, such as cost increases far out-weigh the savings you could realise through moving. But there is an added twist – the cloud lock-in.

All datacentre environments are largely the same – some hardware/infrastructure you own or rent, some virtualisation, some applications and data. But again, largely there isn’t lots of unique services that only that datacentre provides.

In 2023, most of you are probably using lots, if not primarily Platform as a Service (serverless, managed databases, containers). While each cloud may have similar offerings, they are substantially different enough, that moving to an alternative is very difficult, to the point that I would argue the effort and time required to unpick what you have isn't worth it to mitigate the risk or pressure.

Though that’s not to say it isn’t achievable. Depending on your deployment, for example, if it’s Kubernetes based, there are lots of great third-party software platforms that can sit a layer above cloud making this task more accessible. However, it's crucial to consider whether you have implemented such technology from the start or are now considering retrofitting it to manage the risk.

What should you do?

As with most things, my guidance is understanding the risks, together with the short to long term impact, followed by identifying how you can mitigate them. If you’d need to spend considerable amounts of time/effort (translated to cost) planning a move out of cloud, or moving would only realise a marginal saving and your workloads could reside on any platform, with no prevalent cloud benefits required, I would stay put. If the upfront investment is substantially outweighed by a move, in the short and long term, then start planning for your next platform!

Cloud services from Softcat

Work with Softcat for all your cloud technology needs. Our specialists offer an independent view of your cloud journey, offering practical advice tailored to your needs. For more information, get in touch with our team today.