Cloud computing is probably the biggest business around these days — it has become a $100-billion-a-year industry. And there’s a chance companies are paying way too much for it.
That’s the conclusion drawn by Sarah Wang and Martin Casado, partners with Silicon Valley powerhouse Andreessen Horowitz Capital Management, who say the costs of cloud on a business can be staggering. “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,” they wrote in a recent analysis.
Of course, while cost savings in merely the early benefit of cloud, which is presumably subsumed by the flexibility and agility cloud resources provide. Wang and Casado acknowledge as much. “This shift is driven by an incredibly powerful value proposition — infrastructure available immediately, at exactly the scale needed by the business — driving efficiencies both in operations and economics. The cloud also helps cultivate innovation as company resources are freed up to focus on new products and growth.”
The catch is, as companies throw out their on-premises assets and build their applications on top of cloud services, it’s too late by the time they recognize the weighty costs. A rewrite or the significant restructuring needed to dramatically improve efficiency can take years, and is often considered a non-starter,” they state.
Wang and Casado point out that “some companies have taken the dramatic step of ‘repatriating’ the majority of workloads, or in other cases adopting a hybrid approach. Those who have done this have reported significant cost savings.” However, they caution, repatriating applications from the cloud back to on-premises environments can be a difficult and expensive task.
Still, the authors claim that repatriation to on-premises or hybrid environments may results in one-third to one-half the cost of running equivalent workloads in the cloud. “A director of engineering at a large consumer internet company found that public cloud list prices can be 10x to 12x the cost of running one’s own data centers,” they relate. In addition, for large organizations, “the excess cost of cloud weighs heavily on market cap by driving lower profit margins,” further exacerbating these costs. “Across all our conversations with diverse practitioners, the pattern has been remarkably consistent,” they add. “If you’re operating at scale, the cost of cloud can at least double your infrastructure bill.”
When it comes to cloud computing, however, not everything is as straightforward as it seems. When it comes to assessing the costs versus benefits of cloud, the best route depends on a number of factors, says Joe Weinman, author of Cloudonomics and Digital Disciplines. “On the cost side, efficient and competent IT shops at scale with little variability or unpredictability and/or with demanding performance that benefits from a custom architecture can do it themselves,” he mentioned in a tweet responding to the study. However, “with a lack of cost effectiveness due to lack of skills or scale, with highly variable or unpredictable demand, and without a need for a highly specialized architecture due to performance or other drivers, the cloud is a good choice.”
It’s notable, too, that in their analysis, Wang and Casado looked exclusively at the cloud experiences of tech companies, many of whom are cloud service providers themselves, who are voracious consumers of cloud services and data, and also have the on-site skills to effectively run an on-site data center. The question is, do Wang and Casado properly account for the business advancement that is accelerated from using cloud? “As cost isn’t everything: focus on core versus context, time to market, innovation, and revenue growth can also be enabled through public cloud/edge,” Weinman points out.
Granted, it’s hard to quantify such stuff. For example, cloud and SaaS providers embed business process learnings from their range of customers that in turn benefit anyone signing up for a service. In addition, innovators within companies can quickly spin up instances of cloud services to experiment and pilot new ideas. Security is another part of the equation, with service providers staying steps ahead of enterprises — and hackers — in meeting the latest protocols.
The authors offer some suggestions for managing the cost of cloud:
Cloud spend as a KPI. “By tracking cloud spend, the company enables engineers, and not just finance teams, to take ownership of cloud spend…. Developers who have been burned by surprise cloud bills are becoming more savvy and expect more rigor with their team’s approach to cloud spend,” Wang and Casado state.
Incentivize the right behaviors. “Empowering engineers with data from first-class KPIs for infrastructure takes care of awareness, but doesn’t take care of incentives to change the way things are done,” according to Wang and Casado. “A prominent industry CTO told us that at one of his companies, they put in short-term incentives like those used in sales, so that any engineer who saved a certain amount of cloud spend by optimizing or shutting down workloads received a spot bonus.”
Optimize, optimize, optimize. Evaluate the cost of goods sold (COGS) against cloud costs — “for every dollar that a business makes, how many dollars does it cost to deliver? There are a number of third-party optimization tools that can provide quick gains to existing systems, ranging anywhere from 10-40% in our experience observing this space.”
Think about repatriation up front, and repatriate incrementally. “Even modest or more modular architectural investment early on — including architecting to be able to move workloads to the optimal location and not get locked in — reduces the work needed to repatriate workloads in the future. The popularity of Kubernetes and the containerization of software, which makes workloads more portable, was in part a reaction to companies not wanting to be locked into a specific cloud.”