Among the many responsibilities the CEO of a company has, capital allocation is a critical one. In fact, some of the best CEOs have excelled at this function – think Warren Buffet, Jeff Bezos and their ilk. Efficiently allocating the firm’s financial resources to maximize shareholder returns has determined the sustained success of businesses, large and small. As organizations across industries increasingly adopt the Cloud model to run their business technologies and conduct their businesses, efficient capital allocation and capital management has become even more essential. The only difference is that the responsibility is not just with the CEO but must percolate down to every business unit and employee of the firm. In fact, moving to the Cloud model puts the spotlight firmly on capital allocation, monitoring of cost-benefits, and continuous optimization. Let’s take a look at why and how.
The Cloud Advantage
Whether a small startup or a global enterprise, the value offered today to the firm by adopting the Cloud model is undeniable, with the model encompassing private, public, and hybrid variants. Amazon Web Services (AWS), one of the leading cloud providers, in their Cloud Value Benchmarking report of June 2020 point out the key results of their survey of respondents: 27.4% reduction in IT infrastructure spend per user, 67.7% increase in terabytes managed per administrator, 42.4% lower IT spend for applications with over 1,000 users, 56.7% decrease in application downtime, and best of all, 37.1% reduction in time-to-market for new products and services.
I am a big fan of the Cloud model. It has enabled many startups like mine to take off the ground easily to realize our dreams of building and taking to market platforms and SaaS applications that will be used by many firms and users. Companies, small and large, can now rent infrastructure on-demand and easily get access to the latest technologies enabling them to move from capex to opex model, reduce costs, make their business technology more agile, resilient, global, and be able to infuse innovation at scale much more than what was possible before.
A significant implication of moving to the Cloud model is that it is making the organizations much more focused on the economics of running and changing the business – essentially, how much financial and organizational resources to allocate to what aspect of operations and technology, and the expected return on investment. From a Cloud spend perspective, the Cloud economics focuses on how to allocate resources to various aspects of the Cloud and the associated applications, track the total cost of ownership and the return on investment, and continuously optimize to ensure alignment with business objectives.
While moving to the Cloud and operating on the Cloud definitely deliver cost savings, for the cost savings to remain substantial and worth the effort of the move to the Cloud, it is imperative that the business at each level is cognizant of the Cloud economics and take appropriate decisions in every activity as related to operations and technology.
What is also important is to be selective in deciding which workloads to move to the Cloud (or develop on the Cloud) and which ones to continue on-premise. Not paying attention to the Cloud economics comes back and bites businesses especially as they scale up. You want to operate in the goldilocks zone where the economy of scale of operating on the Cloud continues to make sense. Many large firms which operate in a hybrid model (on-premise plus public cloud) have found that when the business has scaled up to a significant level, the same economies of scale achieved on the public cloud can be achieved also on-premise. Then the margins of the Cloud provider show up as the difference. If we are talking millions of dollars in Cloud spend, this Cloud provider margin then works out to be a very significant number. This is what firms like Dropbox went through, this particular story receiving a lot of media attention as an example of Cloud repatriation (moving workloads off from public cloud back to on-premise) way back in 2016.
Another key aspect is, if you are just shifting inefficiencies from on-premise to the Cloud and don’t address them continuously, as your business scales up these inefficiencies also scale up and hit back in the form of exorbitant Cloud bills negating the advantages of moving to the Cloud.
From my own experience as a startup founder, one of the big cost items I found bloating up my Cloud bill was the data-transfer-out cost (also called egress). The more data your application sends out via the public Internet the more egress costs you incur, and in the SaaS model this is what the application does pretty much all the time. It is critical to get a handle on all the unit costs and their drivers at a granular level to first understand where these costs come from and then be able to take sustained action to optimize costs.
So how to ensure you take advantage of the strengths of the Cloud model while also ensuring the economics continue to make sense?
We know how DevSecOps has become essential to the modern agile application development context and delivers clear benefits to predictable software delivery delivered quickly and with quality. This is achieved through automating the build, test, deliver, deploy cycle using infrastructure as code, CI/CD, and containers. Leading organizations are adding to this Financial Operations or FinOps, which is essentially the financial management of various elements and activities integrated into the software development life cycle and operational processes. While some organizations have centralized this into a FinOps function, it is extremely important to embed financial management practices into every level of Ops and Tech, in order to make the Cloud economics continue to work for you.
There are many levers available to make this happen. The illustration below depicts some of the key levers and best practices. While you could have a central FinOps group operating like a CoE, it is important to make sure they are tightly integrated with the Ops and IT teams as many best practices emanate from what individual teams accomplish which can then be disseminated more broadly through the CoE and standardized across the organization.
It is important to understand that all these activities have to be undertaken continuously as the business, technology, and ecosystem evolve.
Questions, thoughts, or need help in your organization? Feel free to message me on LinkedIn.
Also see this week’s challenge here, a short case story on cloud cost optimization.