No one gets fired for cutting costs by 6%.
Until someone figures out that 60% was an option.
In some ways, cloud optimization tools offer perfect cover for IT managers looking to get a good-enough annual performance review. They can bring in software — or hire people — to find single-digit savings on their public cloud bills. And, given the size of some public cloud bills these days, that single-digit is more than enough to justify a bonus.
Provided, of course, no one asks why you’re paying those cloud bills in the first place.
Hammer, meet nail. And screw. And bolt. And wingnut.
Not only do public cloud optimizations earn minor praise with minimal risk, they’re self-perpetuating. Once you’ve found the software or contracted with a firm or even hired an entire team of people to keep an eye on your public cloud utilization, you have a built-in strategy to reduce costs on a regular basis — even if you’re putting that money back into finding more optimizations. You optimize, you grow, you optimize again. Nothing ever gets solved, but it also doesn’t get worse.
The downside of this perpetual pruning approach is that it also never gets better. When you assume that the only tool for the job is a hammer, you end up turning everything into a nail. The same goes for public cloud. It is a near-miraculous technology for the things it’s good at, but that doesn’t make it the best option for what you have.
So many companies come to us halfway through the realization that they don’t have to accept cloud as the only solution for corporate infrastructure. Sometimes, like in the case of 37signals, they’re paying a premium for the option to burst workloads that haven’t budged in years. For others like Picsart, it’s seeing the eye-popping total of running certain workloads in the public cloud — in their case graphical user interfaces, or GUIs.
Ask the real questions
For 37signals and Picsart, taking the time to ask the underlying question — not just “How can I improve my public cloud workloads,” but “Is public cloud the right place for me to be at all?” — saved the companies quite literally millions of dollars.
We get why some people don’t want to ask the question! If you find out that public cloud is no longer the best place for all of your workloads, that can kick off years of discovery, planning and migration. It’s high-stress, high-visibility work, and while it may be highly rewarding, it comes with high risk if you choose the wrong partner.
For some, questioning public cloud can be extra tricky, because they may have been the person advocating for its use in the first place. If we make one thing clear, let it be this: A reassessment today doesn’t mean a mistake was made yesterday. We have been in this business long enough to see the landscape completely change. Not only are the tools for setting up infrastructure different, the businesses that infrastructures have to support are completely different. When we first started working with 37signals more than 10 years ago, we didn’t question them about keeping the bulk of their workloads in public cloud. It made sense for the business and the web that existed at the time.
Embrace the temporary suck of digital transformation
Today, though, needs are different. The tech industry has matured. Many businesses have technical debt causing actual debt in the public cloud. They’ve seen spiky usage level out as their user base stabilizes. Or, they’ve started to rely on workloads that were always cost-prohibitive in the public cloud (our old friend, GUIs).
In those instances and others, it’s time to consider a move to on-prem or managed servers — with a connection to the public cloud for when it makes sense. The discovery process is in-depth, certainly. But doing a deep dive IT optimization review is bound to turn up all kinds of opportunities for improvement that a surface-level scan would never find.
It’s also a chance to step back and think strategically about how your business runs, and how you want it to grow going forward. Your infrastructure should be an accelerant, not a limit. When you get your arms around the business that exists today and the businesses that might exist in the future, you can start to approach IT as a scaffolding to build on.
Escape the status quo and show real value
At Deft, we know our biggest competitor isn’t another firm, or even the cloud hyperscalers. It’s the status quo. There’s a certain school of thought that says it’s better to stick with what you have, making incremental improvements instead of a total overhaul. That can be just fine when your infrastructure only needs the equivalent of a new coat of paint. If the foundations are shaky, though, you could end up doing a lot of surface-level work to a system that’s headed toward collapse.
Talk to us when you’re ready. Together, we’ll build you a better solution from the ground up.
