Cloud Cost Management: How Much is That Puppy in the Window?
We were recently asked by the press to address the question: What’s the most effective method that can be used to gain control over cloud spend?
Our take: Assessing Cloud Costs Requires Context and Perspective.
Cloud spend must be managed within the context of the entire IT cost ecosystem. And the costs need to be put in the perspective of KPIs that identify performance requirements such as business outcomes, user experience, compliance, IT strategy, and tangential impacts.
The most effective approach is to use a dynamic performance model that collects and updates all the operating (OPEX) and investment costs (CAPEX) of an IT/Business system. Ideally, that model starts with a baseline cost assessment of what is currently in place, which is then forecasted in an as-is scenario. If a baseline is not available, performance can be mapped independently.
Costs need context and perspective. This is the dynamic part of the model. Cost is best managed when it is compared to something else, even if that something else is another cost. That other cost could be the as-is cost of doing the same thing or it can be a non-cost metric such as user satisfaction. This head to head comparison becomes the management tool to assess how costs perform in relation to business requirements.
Why is the approach so effective?
One thing we have learned in IT Financial Management is that cost alone is a red herring. A cost must be put into perspective as a ratio of a performance metric. Performance metrics are a function of the business requirements. Business requirements are what make an enterprise unique. Any one-size-fits-all approach will lead to false results.
Follow-up questions from the reporter:
Q. How can organizations plan a successful cloud spend strategy?
A. Model the IT cost ecosystem and forecast. Measure. Communicate. Learn. Communicate. Act. Communicate. Refine. Communicate.
Q. What’s the biggest mistake organizations make when planning their cloud spend?
A. Thinking that outsourcing bad implementation will fix bad implementations. Ignoring and not factoring in vendor lock-in. False expectations (i.e. Cloud is always cheaper). Not factoring in new costs for governance, security, renewals, add-ons, support, customizations, scaling usage and storage, shelfware (yes it exists in subscription models too), etc.
Q. How can organizations lower cloud spend waste?
A. First, define waste. We submit that waste is an excess cost beyond meeting nominal KPI specifications. Note: Every enterprise must define waste within the context of their business requirements and technical strategy.
Second, reduce complexity. The complexity introduced with non-cloud native applications and code, duplicative systems, oversubscribing, multiple buying centers, and other factors can dramatically increase costs. Standardization and consolidation will almost always reduce costs. The modern enterprise is composed of business processes and tools that are often tailored for the use cases (legal, manufacturing, sales, finance, IT, et al). Further, enterprise assets are following the industry trends of abstraction, democratization, availability, and interactivity. These have increased the need for governance and accountability as the visibility of costs has been reduced. However, without the ability to drill down on these cost elements and put them in the context of KPIs, the real root causes will remain hidden.
Third, apply cost management tools. These tools provide a big picture of cloud spend across multiple vendors and work toward eliminating unneeded instances, creeping cost elements, vendor sprawl, unused subscriptions, etc. Automating manual labor, analyze cloud performance and usage dashboards and data, reducing management by spreadsheets can significantly reduce the cost of cost management.
Fourth, play the long game. Create scenarios longer than the annual budget cycle. This will help determine if increasing investment today will lower costs tomorrow. It will also help determine what the impact of budget-cutting today will be on future performance metrics. Further, it will help assess the impact of long-term contracts on cost and savings.
The last thing you want to do is negotiate with the vendors (and increasingly users). Not that you want to skip this fun activity, but it just that negotiation needs to occur once the cost model is created, run through a few cost cycles, ensure the cost dynamics are understood and that a few future scenarios are laid out. Only then should you enter data-driven, fact-based negotiations with your vendors.
Managing cloud spend is an IT Financial Management discipline. However, most organizations do not have a robust ITFM function capable of developing and managing a dynamic cost model – essentially a digital twin of the IT/Business system. Most ITFM software tools do not have the capabilities to create an interactive model. However, the data is largely available, and building an enterprise cost model noted above is 100% possible.
I submit that building this multi-use cost model first, (not a trivial undertaking, but not a terrible one – figure $50-100K of effort), will pay off rapidly by identifying costs that do not add value and eliminating them, without impacting other performance metrics. A note of caution on vendor delivered cost models: They generally only look at a sub-section of the IT cost picture (that which is relevant to what they are selling), so they will be incomplete and possibly misleading regarding the big picture.
I will close by saying that cost and value are in the eyes of the stakeholder. Given that, any cost model should have many ways to dice and slice costs to show the various stakeholders – CFO, CIO, Procurement, business management, IT, etc. Additionally, these cost views should be able to pivot on workload, assets, resources, and process.
The bottom line is that cloud costs are managed as IT costs should always be managed – with a dynamic model, common sense, a holistic approach, and a propensity to over-communicate.
Disclosure: The author is the Founder and Managing Partner at the International Institute of IT Economics (IIIE). The IIIE develops cost and value analysis tools for IT buyers and vendors. For further info go to iiievalue.com