Isaac Douglas, Chief Revenue Officer at Servers.com, outlines why, despite the dominance of hyperscale operations, exercising caution is essential in refining cloud strategies to avoid the allure of superficial and untailored solutions.
For over a decade, the big three hyperscale players – Amazon Web Services (AWS), Google Cloud Platform (GCP) and Microsoft Azure – have dominated the cloud landscape. Now, slowly but surely, the tide is turning.
Key figures and businesses across a range of industries are publicly questioning the ease and financial sustainability of staying in hyperscale cloud.
Tired of paying excessive fees for services they barely use, frustrated by support systems that often seem more like an afterthought than a lifeline and trapped by an ecosystem designed to make it difficult to escape, the conversation around cloud exit is ramping up.
In fact, 81% of IT leaders have been directed by their executives to either reduce their hyperscale cloud costs or halt any additional cloud spending.
While this could be interpreted as a mere cost-cutting exercise, it also signals broader thinking around the use of hyperscale cloud. Namely, that the promised infinite scalability and efficiency of the cloud is not all it’s cracked up to be, and often comes with a large price tag.
To be clear, hyperscale cloud providers like AWS, GCP and Azure do have their place in some organisations. Companies like Netflix and Amazon that experience wildly unpredictable spikes in demand need platforms that offer unparalleled flexibility. Or companies that need burst capacity for peak periods. But for most businesses such extreme fluctuations in demand are unlikely, rendering hyperscale cloud largely unnecessary. At least in the way that it has been used over the last decade.
The rise of hyperscale cloud
It’s important to understand where the cloud came from to compare where it is today and why many companies have been caught up in the hyperscale cloud hype.
Early hosting options from the early 2000s were inflexible and their scalability was low, largely due to long-term contracts and the slow delivery times of additional infrastructure. However, this type of hosting was still better than a business managing its own infrastructure.
When Amazon decided to develop its own tool and server capabilities, it came into possession of a huge set of capabilities that could support the quick scale-up and down of its e-commerce arm for peak events like Black Friday. It was an infrastructure goldmine that Amazon naturally went on to monetise by selling to other businesses, evolving into the AWS brand that we know today.
In those early days, no one else in the hosting market could come close to the capabilities of AWS. And by the time other hosting companies like GCP and Azure built their own clouds, AWS had already established a huge service offering. With these solutions, businesses no longer had to think about compute or storage. All these capabilities were easily available as-a-Service, changing the face of IT infrastructure forever.
But here lies exactly the problem – purchasing hyperscale cloud seems easy enough, but buying the right solution for your business isn’t, and costs can quickly spiral.
The complexity and scaling conundrum
Today, not only is ‘cloud’ synonymous with AWS, GCP and Azure, but it is also incredibly complex. The more services these hyperscale cloud providers introduce, the harder it is for customers to successfully navigate. AWS is now so complex that it even sells courses on how to manage the environment, and companies are hiring people with AWS certifications to help them make the best buying decisions.
The sheer volume of options makes it nearly impossible to determine the most cost-effective solutions without specialised knowledge or costly third-party assistance. This complexity often leads companies to spend far more than necessary, which is not helped by the impossible-to-decipher monthly bills. Thousands of lines in an Excel sheet that take hours, if not days, to figure out. Some companies even end up building internal tools specifically to translate the bills into something understandable.
The investor’s influence
But who is responsible for these poorly navigated cloud purchasing decisions? We can’t overlook the role of investors when it comes to the adoption – and dominance – of hyperscale cloud. Investors, banking on the success of the companies they fund, often push for cloud adoption to ensure scalability, sometimes even naively with the thought that ‘one solution fits all’.
Consequently, it’s not uncommon for investors to overlook the long-term financial implications of hyperscale cloud investments, particularly if the company does not achieve the expected success. This can lead to a vicious cycle of technical debt, where companies become so vendor-locked in their cloud platform that escaping it requires a costly and time-consuming overhaul of their systems.
As a result, we’re seeing far more businesses wise up to the wishes of investors when it comes to investment in hyperscale cloud, instead working in close collaboration to find an infrastructure solution better suited to their business needs.
Proceed with caution
Despite the drawbacks, hyperscale cloud offers many benefits, but any prospective hyperscale cloud solution needs to be fully assessed and carefully navigated. If businesses choose to take advantage of the free credits that the hyperscalers offer, they need to do so with an understanding that what’s free today may come with significant costs down the line.
For any business wanting to avoid or escape the hyperscale cloud trap, an important strategy is to plan an ‘exit’ right from the beginning. Stakeholders should allocate a budget for escaping any technical debt from vendor lock-in, and set firm deadlines for transitioning off a platform before free credits run out.
While it may not always be possible, developing a vendor-agnostic cloud infrastructure is one of the safest avenues for businesses looking to expand their capabilities. Building a solution that is not tied to a single provider will save time and money in the long run, because untangling an organisation from one provider in favour of another takes both time and investment. With hyperscale cloud, it isn’t just a simple switch, but a whole re-architecture of systems.
There is still much to be done about the reliance on hyperscale cloud, but the growing conversation around its complexity and sustainability is a positive step forward. More businesses are re-evaluating their long-term cloud strategies, fully considering the wealth of options now available to them. Are the glory days of hyperscale cloud dominance over? Only time will likely tell but industry conversations suggest that at the very least, the shine is starting to come off.
