Don't blame the cloud if you're failing to continuously improve

This short post in response to an article I read from a16z https://a16z.com/2021/05/27/cost-of-cloud-paradox-market-cap-cloud-lifecycle-scale-growth-repatriation-optimization/ which says

While back-of-the-envelope analyses like these are never perfect, the directional findings are clear: market capitalizations of scale public software companies are weighed down by cloud costs, and by hundreds of billions of dollars. If we expand to the broader universe of enterprise software and consumer internet companies, this number is likely over $500B — assuming 50% of overall cloud spend is consumed by scale technology companies that stand to benefit from cloud repatriation.

What's analyzed and illustrated with the of Dropbox - is a perfect illustration of how finance and investors are looking at companies, with a touch of sensationalism in the title.

It's easy to blame public cloud providers for costing a lot to no longer fast-growing companies. And yes, guess what: when a company plateaus or goes down, the natural financial reflex is to cut the costs to improve margins. I feel angry when I read one of their conclusive statement

"You’re crazy if you don’t start in the cloud; you’re crazy if you stay on it."

Public Cloud is a service, hence you pay for a service and you should always look at it this way. You don't pay for a playground where you create a mess and don't pay attention to fundamental aspects of your product, architecture, and operational aspects. I'd argue that any company that "let go" on good architectural principles and discipline is a failure-in-the-baking company. Sooner or later, it bites. And yes, like Dropbox, they had to cut costs rapidly to show their investors a quick win. Repatriation might work as it's an upfront investment that can be amortized, and you stop having recurring expenses. You think.

In reality, if a company let cloud expenses grow without control (that is a business grows), then they have no control over their technical debt, and sooner or later they will get stuck. A good counter-example to Dropbox is Netflix. Afaik, they are still in the cloud and they have learned to observe, redesign, implement, improve, etc. To a point that they even donated to the public domain (open source) their micro-services infrastructure. Why? because they went beyond that as a result of constant observations and improvements.

So a16z goes with the fallacy that "it’s becoming evident that while cloud clearly delivers on its promise early on in a company’s journey, the pressure it puts on margins can start to outweigh the benefits, as a company scales and growth slows. " It's a short-sighted perspective. Early in a company's journey, your build a product on an architecture that is good or bad. The less you care about your architectural design, the more stuck you will be over time. Refactoring, clean code, well-architected principles are not eroding margins. Lacking this discipline erodes margins. For the most part companies get eventually stuck with deployment models that don't scale on cheap (cloud) resources. They scale vertically only (well, there is more to say, but trying to keep this short), and yes, cost increases in public cloud. A solid and flexible architecture from day one will let you use one or more cloud providers as required and consume what's required. Of course, nothing prevents you to have a hybrid deployment (on-prem and cloud), especially if all cannot be dematerialized (physical presence). But you can't blame the cloud!

A don't forget that once you're back to your private cloud/data center after a 'repatriation', I bet that other costs will surface. Think security for example... oops, who is going to validate that all software is patched on time? disaster recovery in separate geographical regions. Etc.