Learn how to use unit economics to understand the relationship between your business growth and cloud costs. Improve your profitability and sustainable growth with Cloudthread's cloud cost unit economics.
Profitability and sustainable growth are top of mind for all businesses in today’s economic climate. In the cloud, this means understanding the relationship between the growth of your business and the growth of your cloud costs to serve your customers. Many companies have been operating with a blank check mentality around cloud costs and don’t know what this relationship is. They don’t know what their cloud cost of revenue is and they don’t know how business growth forecasts will influence infrastructure costs. This is where cloud cost unit economics help.
In this article, we will discuss how to construct a unit metric, what cost of goods sold (COGS) is, how it relates to cloud costs, and how businesses can use unit economics to forecast their cloud costs.
For the purpose of this article we’re referring to business unit metrics, as opposed to engineering unit metrics - for more context on definitions see here.
Fundamentally, a numerator and a denominator. Typically the numerator is a relevant segment of your infrastructure costs and the denominator is a metric relevant to a unit of demand relevant to your business.
What makes segmentation of the numerator tricky is that you want to isolate infrastructure that’s relevant to the unit of demand you’ve selected and the variable cost to serve that unit of demand. You don’t want R&D spend included. You don’t want infrastructure dedicated to testing and development included. You don’t want fixed cloud costs like support included. Ideally you have adequate tagging or account segmentation to isolate production costs relevant to the product/org/team you’re looking to isolate unit metrics for.
For a SaaS company with a single product and a single pricing tier, this could be as simple as isolating production costs.
For a SaaS company with a single product and multiple pricing tiers, this could mean breaking out different unit metrics and segmenting infrastructure for each pricing tier.
For a SaaS company with multiple products, numerator segmentation means breaking out infrastructure spend associated with each product.
This is very business specific. It boils down to what the unit of demand for your company/product is and ideally also ties directly to how you bill your customers.
At Cloudthread, our unit metric is $ / customer and we distinguish between free tier users, startups, and enterprise customers. Clever wants to understand their $ / login. Lyft wants to understand their $ / ride. For one of their products, Stedi wants to understand their $ / M API requests. SaaS companies want to understand their $ / user / pricing tier across products. An eCommerce company wants to understand $ / order.
Cloudthread’s unit metric features provides a cloud cost data pipeline to slice and dice the numerator in order to isolate relevant infrastructure and a flexible data ingestion API in order to ingest denominator metrics that are relevant to company products. Unit metrics are then the foundation for benchmarking, reporting, anomaly detection, and custom dashboards for leadership.
In the future, we’ll be launching features focused on scenario modeling by tweaking infrastructure costs (e.g. Product X AWS cost) and demand driver variables (e.g. # of customers).
.png)
Cost of goods sold (COGS) is the cost of producing and delivering a product or service. For a company selling phone cases that cost $1 to make, their COGS is $1. In the cloud, COGS is the variable cost of infrastructure to deliver your product to customers. To understand your COGS, you need to calculate the infrastructure cost of each unit of your product or service.
To calculate the cost of each unit, you need to break down your costs into fixed and variable costs. Fixed costs are those that do not change with the volume of units produced, such as enterprise support or infrastructure costs to support internal teams. Variable costs are those that change with the volume of units produced for your customers. Typically you want to exclude engineering costs associated with R&D, development accounts, and testing.
Engineering teams are constantly asked to create budgets around cloud costs.
At companies with no unit economics and no collaboration between engineering and finance, this typically looks like an engineering manager hacking estimates together that are unrelated to business growth forecasts with the AWS calculator, and giving an estimate that will almost definitely be off.
At companies with a unit economics framework to analyze infrastructure costs, engineering teams understand the relationship between the growth of their business and the growth of their cloud costs to serve your customers. Forecasts are informed by extrapolating the unit metric variable cost (e.g. $ / user) into the future based on business growth projections and layering on any relevant fixed costs.
In this article we covered what makes a unit metric and how unit economics are the foundation for understanding your cloud COGS and cost forecasts.
If you’re looking to adopt cloud cost unit economics at your organization and could use support, please reach out and we’d be happy to help!
