The cloud world of self-service procurement and highly variable consumption presents a new set of challenges for TBM's promise of delivering value from, and properly allocating, IT expenditures.
Cloud spending now accounts for a growing portion of IT spend. Yet with all this growth, the cloud operating model is still immature.
A new operating model for the cloud, called FinOps, has emerged, enabling a shift in how cloud is managed.
In the same way that DevOps revolutionized development by breaking down silos and increasing agility, FinOps increases the business value of cloud by bringing together technology, business, and finance professionals with a new set of processes.
In the following recording from TBM Conference 2019, JR Storment and Mike Fuller from the FinOps Foundation look at how TBM and FinOps work together to provide a complete framework for making business decisions about cloud spending.
For those who prefer a written version, Apptio has also provided a transcript after the video.
Everyone is trying to derive value from cloud, and cost isn’t the best (or the only) metric to quantify value.
DevOps and Cloud has broken traditional procurement models.
Engineers now spend company money at will—every minute of every day. Engineers innovate at high-speed and spend at high-speed.
Companies are driving their business with software. As a CIO of a major US airline says, “We are no longer an airline, we are a software company with wings.”
Moving to a cloud-centric world isn’t merely an accounting change from CapEx to OpEx. It’s a change from a fixed to a variable view.
Traditional IT consumption
For decades, the IT consumption story has remained the same.
The players (engineers, finance) operate in their own siloes—engineers request, finance approves. Platform team make plans and then talk to procurement. The model is slow and approval-heavy.
Elongated procurement cycle, adding a delay between identifying a need and delivering a resource, increases the odds that the output (when finally delivered) is no longer needed or relevant.
The cloud-centric model
With DevOps, you build it and you run it. No-one is getting in the way. Platform teams have dissolved into Service teams who need capacity to run their services.
When they move to the cloud, automated processes (scripts, auto-scaling, CI/CD) ask for capacity—not people. What’s a win for innovation and a loss of control for Finance. A cloud-centric model, without FinOps, promotes a lack of communication between engineers and Finance.
New Cloud realities for IT
- Decentralized—buying centers for cloud are siloed from Finance and central IT.
- Material—24% YoY increase in public cloud spend
- Variable—variable/communication-based cloud spend replacing datacenter/fixed cost spend
- Inefficiency—macroeconomic instability pressuring cloud spend and usage towards efficiency
FinOps brings financial accountability to variable spend with:
- A prescriptive model of actions, best practices, and culture—designed specifically for cloud.
- A common lexicon and methodology for product, finance, and business teams.
- Decision making that increases business value.
FinOps isn’t about cutting cloud spend. It’s about putting the right people and processes in place to make on-going intelligent decisions.
What’s in this new world?
Engineers get to work fast and innovate, but also want controls and accountability around their cloud spend to ensure maximum value from spend.
There is a slice of FinOps that involves cost optimization, but it’s only a part of it.
It’s about the collaboration between finance and engineers: Educating teams to work together, using appropriate reporting, setting goals, and using automation to make FinOps work at the speed of cloud.
FinOps has a life-cycle. Less of a finite start-stop process and more of a flywheel that spins faster and faster in an organization.
- The inform phase takes a measure of what you are spending today, and where your cloud spend is going.
- The optimize phase looks for efficiencies that meet goals while balancing the quality and cost of our products.
- The operate phase is where the action happens: workflow, processes, accountability, and automation.
The cycle then starts again at the inform phase to remeasure spend.
Crawl, walk, run: achievable goals and the FinOps lifecycle
Start with small, achievable goals and then learn from them throughout the lifecycle.
As the lifecycle iteration build, organization build confidence and a track record of success—moving from crawling to walk, and then run.
Successful FinOps practitioners end up looking at unit economics (e.g., cost per customer, cost per transaction). Basically, looking at the unit economics of the “thing” that’s driving revenue.
FinOps and TBM are complementary
While FinOps has great controls into cloud, granularity, and prescriptive methods, it's missing TBM data.
FinOps focuses on cloud processes, cloud data, and cloud execution: TBM brings in the general ledger (GL) data (labor costs, licensing, revenue data).
Together, these deliver a unit economic cost that is fully burdened, accurate, and can be allocated to the right places.
The development teams, the innovation engine in your organization, are unblocked and can work at speed without losing control of costs.
The three “legs” (FinOps, TBM, and development teams) stand up the cloud economy. Start with granular cloud data coming out of FinOps and labor, cost, and revenue from a GL. Together, they deliver an accurate cloud-chargeback. If you are doing cloud at scale, without FinOps processes in place, chargeback is probably inaccurate.
Cloud chargeback delivers unit economics.
TBM surfaces larger business decisions and a methodology behind what to do next. That data feeds into the processes of FinOps. Cost optimization may be one of the processes, but it could also mean spending more, or it could be about changing how you run databases by adopting RDS or Aurora.
All works toward the final point of increasing business value.
TBM takes a top-down view; FinOps from the bottom up. TBM looks to the fully-burdened all-up view of IT spend; FinOps looks at hourly costing and identifies infrastructure anomalies in real-time
Pre-purchase the FinOps Foundation book (authored by JR and Mike)
Cloud FinOps O'Reilly