I’ve been fortunate in my career. Like many, I’ve witnessed the evolution of my field multiple times—from mainframes, virtualisation, and monolithic architecture to DevOps, agile, automation, and now AI. The future often seems to point toward machines ruling machines. But despite the advancements, we’re still distant from this sci-fi reality.
Today, the cloud appears to be the ideal solution for any reputable company to embrace. After all, if major tech companies and competitors are using it, shouldn’t I be using it as well? It’s a reasonable assumption, but it comes with its trade-offs.
I believe that being on-prem or having your data centre is still an optimal decision, and if you do your maths right, you will save vast amounts of money.
But don’t ignore what cloud computing can do for you, especially the public cloud. This is where the don’t repeat yourself is very handy.
Culture is Key
The first trade-off is cultural. We all know that culture shapes a company’s competitiveness. Tools are essential, but in the wrong hands, they’re as helpful as a boat made of stone. It is moving to the cloud “just because” is a misguided approach. If you plan to perform a lift-and-shift migration and then innovate, you may be in for a nasty surprise: unexpected costs, a frustrated team, and a host of new daily incidents. While starting from scratch with a cloud-native architecture might seem daunting, it can be better than trying to patch up a system not designed for the cloud.
Throughout my career, I’ve approached cloud migration cautiously. I plan to start with the simplest, non-critical systems, beginning with the database. This allows me to see how connectivity behaves between on-premises and cloud databases and gradually move the system to the cloud. But by “move,” I don’t mean a mere lift-and-shift. I mean rebuilding the system in the cloud, connecting it with the backend, and gradually transitioning from on-prem to the cloud setup. If something goes wrong, we can revert to on-prem and reassess the issues.
This cautious, phased approach builds confidence and peace of mind, ensuring the team has a fallback plan rather than dealing with extended periods of firefighting.
And leadership needs to be educated in this way. I suggest simply putting current systems in maintenance mode and ensuring they run the business with minimal new features. In my experience, features are often nothing more than quick bug fixes for what is already damaged. They’re essentially damage control to cover up someone’s mistakes or to maintain a particular status quo.
Don’t be afraid to suggest pausing some development. You’ll likely see the number of incidents decrease significantly. Trust me—I’ve seen it firsthand, and interestingly, incidents are fewer during holidays. But that’s a story for another time.
“The Cloud Will Save Us Money” – The Myth and Reality
When I hear declarations like, “The cloud is the best; it will save us money,” I feel compelled to address some uncomfortable truths. Yes, the cloud has many advantages, but it has significant trade-offs. The solution lies in balancing the pros and cons and approaching migration as a transition rather than a quick fix.
No one ever seems to have the time to plan a thoughtful migration—until a major system meltdown occurs, and someone demands a return to on-prem no matter the cost. Moving to the cloud requires education, cultural preparation, and training. A successful migration means selecting a team to experiment with the cloud, complete a few projects, and provide feedback to train others. It’s a cycle of continuous improvement.
Relying solely on job titles like “DevOps Engineer” or “Cloud Engineer” to drive transformation is more about looking good in a presentation than achieving actual cloud-native status. Becoming genuinely agile or cloud-native is a journey that takes years, not a hiring spree.
The Reality of Costs and Security Concerns
As for saving money, it’s possible, but it requires tweaking and sometimes swallowing your pride. Managing cloud costs effectively often demands an operational overhaul and vigilance that some organisations find challenging. The cloud also means entrusting your most sensitive data to an external provider, which raises security concerns. Your “crown jewels” are essentially in someone else’s data centre. While cloud providers invest heavily in security, companies must be aware of the risks associated with storing critical data outside their control.
In my experience, the cloud works best for temporary workloads—tasks that can be scaled up and down quickly or temporarily expanded as demand requires. I prefer an on-premises setup for mission-critical systems, using the cloud for disaster recovery or additional resources during peak times.
Don’t be Like This
The worst thing a business can do is drop the cloud on top of engineers, especially if they aren’t proficient with it. Expecting them to learn quickly while handling frequent incidents, juggling 30 to 90 projects per quarter, and receiving no roadmap or guidance from leadership is a clear sign of a toxic workplace and an inefficient business model.
Educate your team, involve them in preparing for new initiatives, and introduce new ways of working. This approach fosters confidence, creates alignment, and ultimately leads to more sustainable success.
And if you are part of leadership and don’t follow through and believe all is fine, be prepared for multiple incidents and possibly an exodus of engineers, not to talk about losing the trust of your business partners.
The Cloud as a Learning Opportunity
The cloud is also an educational resource, demonstrating how large-scale data centres are managed. AWS, Azure, and Google Cloud Platform have perfected complex, resilient configurations across multiple regions. By studying their methods, companies can aspire to achieve similar setups on a smaller scale, though this requires significant investment in time, effort, and training. And while outsourcing everything to the cloud might seem more straightforward, there’s value in building in-house expertise. It reflects a modern approach to business—operationally light but intense in response and adaptability.
Will the Cloud Be the Future?
I believe so, but it’s essential to understand that the cloud is the ultimate vendor lock-in. I envision a future where companies use the cloud primarily for scalability and flexibility—connecting, scaling, and deploying resources on demand while keeping their core operations on-prem. Imagine a setup where a company’s entire infrastructure fits in a compact, highly efficient data centre. Picture running a business from a data centre the size of a Raspberry Pi—one that’s easy to duplicate, relocate, and rebuild as effortlessly as buying a mobile phone. That would be revolutionary.
Providers will always claim their solution is the best; we’ve seen this with mainframes, virtualisation, and containers, each adding more complexity than before. It’s wise to educate yourself and not be swayed by trends. FOMO can be elusive and persuasive, but you may avoid serious mistakes if you wait. Providers must promote extensive benefits because their goal is profit. I remember when a one-week VMware training course could cost between $4,000 and $6,000, and even the manuals were expensive, and now, well we all know how things with VMWare are.
Too Much Cloud Documentation
Today, when someone says they can’t train on the cloud, I ask if they’re using search tools effectively. All major providers offer free cloud access for a limited time, and the credits they provide are usually enough to build confidence in training. Plus, there’s YouTube and other learning platforms. Cloud providers don’t offer extensive documentation out of courtesy; it serves them well—the more people know about their product, the better it sells. Back in the day, having a CCNA or VCP was considered elite. Today, there’s no such luxury, as competition is fierce.
Final Thoughts
The cloud is a powerful tool, but it’s not a cure-all. Moving to the cloud requires more than technical expertise; it demands cultural alignment, strategic planning, and a willingness to rethink old working methods. Embracing the cloud wisely means assessing your company’s unique needs and finding a balanced approach that makes sense. By carefully planning the transition and building in-house skills, companies can use the cloud as an enabler of innovation without sacrificing control or oversight.
Ultimately, the cloud may be the future, but how we use it today will define whether that future is prosperous.