Financial Operations and Sustainability

While the optimal FinOps strategy for every organization is different, successful FinOps implementations will always require a holistic mix of people, process, and technology considerations. As a result, it is necessary to understand the organization and its way of working before setting up specific FinOps strategies or practices.

Common elements of integrating FinOps with the operating model include collaboration across stakeholders including finance, account, procurement, IT, and business teams. This will enable a continuous lifecycle of understanding Cloud demand, requested Cloud resources, and optimizing Cloud resource use. Additionally, it prevents silos of knowledge which prevents successful FinOps adoption. This is further explored by the FinOps foundation as part of their Teams need to collaborate principle:

"It’s imperative that teams work together to improve the FinOps practice and achieve continuous improvements in efficiency and innovation. Collaboration between cross-functional teams can enable finance to match the speed and granularity of IT, empower engineers to treat cost as they would other efficiency metrics, and help establish standardized governance and controls around cloud usage.

In summary, an integrated operating model which details the standard roles, responsibilities, processes, metrics, tools, governance, and enablement is critical to holistic coverage for FinOps.

FinOps is generally focused on behaviors and processes and is be independent of technology layer – i.e., infrastructure, platform, middleware, and application/services teams can all apply similar requirements to gain efficiencies from their respective solution areas. However, stakeholders in the organization may not understand their role in making FinOps successful.

By establishing clear FinOps goals such as optimizing or managing Cloud storage costs or communicating specific FinOps target metrics, organizations can derive specific measures or actions which each stakeholder can apply to the work they do.

This principle is broadly part of a goal to create a cost conscious organization that is aware of cost drivers and assets they can use to make better decisions. Achieving the ideal state requires consistently gathering and publishing timely data, as referenced in the FinOps Principle of Reports should be accessible and timely, so that teams are making the correct decisions backed by actual usage data. For more information see the FinOps Foundation Principle of Decisions are driven by business value of cloud.

One of the most common challenges of a FinOps implementation is the alignment of strategies and goals to team behavior. As a result, organizations need to implement processes and guardrails to guide desired FinOps behaviors. For example, establishing ways to encourage and enforce practices such as variable cost management, just in time resource procurement, lean resource management, etc.

This FinOps governance should be defined and agreed upon by all involved teams so that there's no surprises. Note that it's important that governance is not a bottleneck for FinOps, instead they are lanes to guide teams when they are in certain scenarios.

As an example, an organization may develop a policy which provides flexibility for teams to freely select cloud resources to a certain clip level or to those that have been at or below their budget guidelines. In this way, governance rewards appropriate behaviors and helps address the misconception of FinOps being only a means to reduce cloud spend as there may be a genuine reason to select a given cloud instance type, even if it's the not the cheapest option. For more information about individual behavior in the context of FinOps, see the FinOps Foundation principle of Everyone takes ownership for their cloud usage.

In the maturity context of FinOps, a critical end state involves the use of automation to not only reactive to scenarios, but have proactive actioning as well. Organization should strive towards automated optimization based on analytics; for example, usage metrics, spend, utilization, etc. using appropriate tooling.

As defined by the FinOps Foundation Principle of Take advantage of the variable cost model of the cloud. To ensure maximum value from cloud spend, companies must take advantage of cost savings opportunities in the cloud cost model. This includes comparing pricing options and usage discounts offered by various service providers and rightsizing instances and services purchased.

This is difficult to complete via manual methods so organization should adopt an automation bias as new FinOps practices are piloted and scaled. Automation may focus on the following elements of a notional FinOps framework:

• Instrumentation: tools to help with data gathering to track and log usage.

• Tagging: tools for accountability of usage.

• Procurement: tools to correct and enforce Cloud specific practices for contracting; for example, contract structures, pricing types, etc.

• Reporting/Invoicing: tools to track, forecast, and/or invoice Cloud costs.

• Solution Efficiency: tools to optimize solution design processes and well architected frameworks to design efficiently.

As part of an organization's existing sustainability strategy there may be specific metrics or results which teams strive to achieve. While some may not impact IT, there are other broader guidelines such as emission reduction targets, efficiency goals, or vendor preferences, etc. which will be relevant to IT and Cloud decisions. Teams should be aware of these goals, but also balance them against critical requirements to ensure that business or workload performance isn't impacted. See examples below:

1. Workload Placement:
   • selecting a data center with lower emissions vs
   • selecting a data center with better latency
2. Efficient Hardware:
   • refreshing components to reduce power consumption vs
   • prolonging component use for better business value
3. Code efficiency
   • running independently scalable code functions for less resource utilization vs
   • running grouped applications due to security and user experience considerations

Acknowledging this difference will help teams evaluate decisions and balance viewpoints which may include competing priorities. It will also create purposeful decision making with reasoning if one choice is emphasized over another.

Given the increased use of pre-developed "as a Service" capabilities like Cloud, organizations must take a proactive view to ensure sustainability by design. This requires applying sustainability goals against the procurement lifecycle to account for requirements from the start. Examples:

1. solution design: establishing efficient designs that encourage reuse and minimal waste.
2. requirements definition: incorporating mandates for meeting certain sustainability standards and certifications; for example, ISO and Sustainable Development Goals.
3. partner / solution selection: prioritizing solutions with committed sustainability targets and achievements.
4. operations: encouraging the use of governance to reduce waste, etc.

A basic starting point for teams seeking to increase sustainability is measuring and track impact. This can be done using various methods, tools, processes, etc., however it is key to establish a baseline as a way to drive sustainability related behaviors and measure progress.

Impact may be tracked via qualitative or quantitive data and align with the strategy objectives defined by the organization. In addition to measuring and benchmarking, this type of data is critical for investment decisions as teams can utilize various metrics to justify resources requirements and forecast future initiatives.

Finally, impact data may have external consequences to an organization. As part of various governmental and non-governmental policies, sustainability data is increasing used to view regulatory compliance and communicate commitment to ESG initiatives. End customers may also use this data to establish partnership decisions.

FinOps strategies, policies, and governance should be driven by a single accountable centralized team. This enables the highest clarity of focus and facilitates data accuracy which is critical for FinOps success. Note: this centralized team will collaborate with numerous responsible parties to enable the appropriate outcomes. It also enables greater synergies which can be used as part of Cloud pricing negotiations.

Deployments should support tagging and account hierarchy to enable better usage allocation and visibility. Tagging is a way to categorize and group instances to enable aggregation of usage across teams, products, business units, etc. For example, mandatory tags to allocate responsible business division for a workload to gain visibility and enable cost optimization for specified domains like HR, Accounting, Engineering, etc.

Account hierarchy refers to the set up of user and instance accounts to mimic your organization's structure. This enables easier mapping and tracking of usage between a given account and the responsible owner.

When combined, tagging and account structures gain visibility into cost per each category in alignment with any budgets and forecasts.

Deployments should be integrated with selected IT cost management tools to enable automated data for cost and usage. Similar to other observability-focused use cases, data availability is not a bottleneck, instead teams are often limited by how to merge financial data across systems, allocate to the appropriate users, and establish FinOps reports. By focusing on automation as part of FinOps startup, tasks can be designed as part of a workflow to make it easier to proof FinOps value and free up teams for more value-added tasks.

One common driver for high Cloud costs is non-production resources spun up for experiments. Additionally, users may select the largest instance of a Cloud service when lower tiers would suffice. To prevent this, implement pre-approved sandbox type environments where users are free to choose from a list of Cloud services and automated reclamation prevents unnecessary uptime.

This helps manage costs while allowing teams to trial infrastructure/platform capabilities and minimize non-value-added billing.

Depending on middleware/application/service type, deployments should be designed with an “as a Service” mindset to enable granularity into cost of usage and reporting.

This is especially key for new services being built with FinOps in mind. Teams should start by identifying what components at each layer are applicable targets for cost optimization. This is determined by being able to tag or allocate specific component services to a distinct accountable party or accounting category as determined by accounting and finance teams. Next, components should have an associated cost element to enable consumption-based pricing.

Standard definitions, terms, metrics, report should be defined and teams should be enabled in the appropriate application of these concepts to produce reliable outcomes. Metrics are especially critical as they enable benchmarking and better decision making.

Organizations should create a cost conscious culture so that each team member understands what FinOps is, how it impacts the organization, and what steps can be taken to achieve the right results. This includes understanding fully loaded costs and making decisions to rightsize resources.

A common starting point is establishing showback, which is a way to make teams aware of costs incurred, and/or chargeback, which requires formally invoicing teams for costs incurred. This is typically accomplished by working with appropriate finance and accounting teams to aggregate costs for a given service and allocate it to the appropriate unit; for example, by resource, team, or user, etc.

As part of broader FinOps strategies, policies, and governance, define and prioritize Sustainability targets for areas such as Platform, Application, Data, Operations, and People & Culture. This allows teams to measure and track progress via quantifiable data and helps establish sustainability-related accountability for stakeholders across all parts of the business since individuals understand what they can influence.

For example:

• Platform and Infrastructure: targets for carbon emissions and environment-related packaging impact
• Application: targets for energy usage
• People: targets for social impact such as talent sourcing

Establish Sustainability viewpoints for Enterprise Architecture teams who are a critical partner for business teams as they design Cloud solutions. They are uniquely placed to coach and influence more efficient solution design with considerations for workload placement.

Some example decisions that Enterprise Architects can influence include:

• selecting data centers with better ESG scores or those with environmental certifications
• ensuring optimal resource selection to not over size instances
• optimizing solution design by refining code to remove unnecessary operations or minimizing processes

Develop and publish a Sustainability dashboard with key metrics that are broadly accessible by teams. Focus on showcasing current status and future goals so teams can pivot as needed to achieve certain goals. Additionally, this is critical for understanding aggregate impact and alignment to Sustainability strategies.

Dashboards can also facilitate reporting which many regulatory and non-governmental organizations request as part of understanding sustainability impact and progress.

Finally, dashboards are helpful to prove value of sustainability initiatives and direct investments to achieve target goals.