How unit metrics help create alignment between business functions

Voiced by Amazon Polly

As the last blog in the Unit Metric series (intro, what is unit metric, selecting a unit metric to support your business, unit metrics in practice – lesson learned), we’ll share how unit metrics is instrumental in gaining alignment across business functions. High quality unit metrics create an opportunity to measure progress toward shared goals. This helps support everyone in the enterprise to move in the same direction and drive toward the same business outcomes. Business functions, such as accounting, product management, pricing, and engineering, individually and collectively, can benefit from well-constructed unit metrics. Before we begin, you may wish to review unit metric basics presented in this blog series.

How does alignment across business functions take place as a result of using unit metrics?  A well-constructed unit metric provides a common point of reference, a “North Star” if you will, that can be used by accounting, financial planning and analysis (FP&A), pricing, product management, sales, and marketing in addition to engineering.  We’d like to take a moment and stress the next point.  With unit metrics and other high quality quantitative key performance indicators (KPIs), you can move toward establishing a data driven decision making culture.  Track changes to your unit metrics and KPIs after feature releases, architectural changes, bug fixes, etc., and let the data inform everyone as to the impact of the change.  Don’t guess.  Did you lower the cost of operations or not?  Did the number of requests processed per hour go up or down?  What was the actual savings achieved by that project to store infrequently accessed data in lower cost storage classes?  Stop guessing.

Let’s begin with a bit of project management terminology. A non-functional requirement (NFR) specifies criteria that are used to determine how system should perform, whereas a functional requirement defines what the system should do.  Establish a NFR for work to be placed in the backlog, an estimate of the expected AWS operating cost.  Assessing and evaluating the cost drivers in a pay-only-for-what-you-use ecosystem is an important and often overlooked element when developing and deploying code.  Modeling the expected operating cost for new code or the expected change in operating cost for existing code will provide an important data point when it comes to prioritizing the backlog.  The operating cost NFR will provide guidance as to the projected impact of those changes to a product’s financials.  This NFR also supports the critical education objective of teaching architects and engineers about the cost drivers for the AWS services they are evaluating, deploying, and operating.  Having an understanding of the cost drivers for an AWS service is as important as knowing the proper use cases for and capabilities of that AWS service.

Whether you are using the AWS Pricing Calculator or building model in an Excel workbook, planning for the cost of operations is key to aligning all of the vested interests in a product or service’s success.

By modeling the operating costs through unit metrics:

• Product owners can better prioritize work that improves gross margin* against work that delivers new or improved features, functions, and capabilities that will either increase market share or reduce churn.

*For non-accountants: adding a dollar of gross margin (net sales revenue) is different than earning an additional revenue dollar.  Talk to your partners in accounting and finance and learn about that difference. We’re not going to tell you here since a key tenant of Cloud Financial Management (CFM) is to build a partnership between Engineering and Finance organizations.  Use this as an opportunity to work on that partnership.

Product owners can use the operating cost NFR as an additional data point when assessing the priority of how work is slotted in the backlog.  Based on our experience, there is a strong bias for new product capabilities taking priority over the remediation of technical debt inclusive of Cost Optimization (CO) effort in a product’s backlog.  Using operating cost or the change to operating cost as a NFR can give CO efforts a better seat at the prioritization table.

•  Marketing can measure the efficacy of a program or campaign by tracking changes to a product’s demand driver (the denominator of a unit metric) against the investment made to the marketing programs or campaigns (the numerator of a unit metric).  Adjustments in investment or prioritization among marketing programs or campaigns can be made in a more objective way.

• Pricing can use the financial variation of a unit metric, the one that presents unit cost, as a means to ensure a product’s target gross margin is appropriate when developing a proposal for a customer. The change (forecast or actual) to a unit metric above a predetermined threshold can also be used as a trigger to revisit a product’s pricing strategy.

• FP&A will perform analysis using the unit metric in a given period to assess if target gross margins (forecast) were achieved for products or services in the company’s portfolio.  Unit metrics will also be a tool to help explain operational cost variances.  By partnering with Product and Engineering, FP&A can utilize a product’s roadmap to forecast where unit costs will be on a go-forward basis — remember that NFR to include operating costs when specifying work to be put in the backlog?  When actual costs are captured, finance can work with their partners in Engineering to explain the variances.

• By using a well thought out structure for AWS accounts, cost allocation tags, AWS Cost Categories, or a combination of these, accounting can assign the actual expense of operating a product or a service to the product’s cost center and reduce the amount of IT spend that is allocated via the “peanut butter” spread method.  Having the option to assign identified costs where possible is always better than having to rely solely on estimates via allocation methodologies.

Leadership can establish cost targets for unit metric KPIs and align incentives to promote the outcomes that are most important to the organization.  Sometimes the business goal is to increase the market share, the demand driver and denominator of a unit metric.  During the COVID-19 pandemic, driving down unproductive cloud spend, the numerator of a unit metric, has become front and center for many customers.

A product’s KPIs, of which a unit metric should be one, is recommended to be published and available on a regular basis via mechanisms such as dashboards or automated reporting to everyone in the company.  Technical and non-technical team members alike should be educated on how each product’s unit metric is constructed, why a given demand driver was selected, and the expectations placed on their teams to move their KPIs in the right direction.  To jumpstart your cost dashboard journey, take a look at the AWS cost management report tools and Well Architected Labs cost pillar.  The web-architected lab provides step by step guidance on how to generate the AWS Cost and Usage Report (CUR) and visualize it in Amazon QuickSight.  The 300 level labs even provide code for a variety of useful queries that will help you parse the CUR.

The combination of learning about the cost drivers for the AWS services being used, well-functioning unit metric KPIs, and open two-way data-driven communications between the engineering, accounting, pricing, marketing, and product teams will help your organization generate more business value from its use of AWS.

We appreciate your taking the time to read this blog series and hope you learn something useful for your Cloud Financial Management journey.  We look forward to hearing about how you identify, implement, and benefit from the use of unit metrics.  If you’d like to share your story about this topic, or any other IT Financial Management topics, feel free to drop us a line at cfm-casestudies@amazon.com.