Mature Companies Don’t Care About Cloud Costs

“Cut spending!” “Slash costs!” “Stick to the budget!”

Poke your head into almost any finance meeting in a SaaS company and you’re likely to hear one or more of the above phrases played on repeat. At first glance it makes sense: Costs are increasing, so we should reduce them.

I’d like to challenge that narrative. Mature companies don’t care about cloud costs.

Don’t get me wrong; I don’t mean to say that once a company has reached the advanced stage in its FinOps journey (which makes it “mature”), it can suddenly throw money away without worrying. What I mean is that it now has the ability to shift its focus from reducing costs to proactively managing them.

This shift marks a company’s evolution from beginner to advanced in the FinOps landscape.

Related read: Granular Allocation, Accurate Unit Costs: The New Standard For FinOps In The Outcome Era

It also transforms cloud expenses from inescapable overhead that must be reduced to a lever that can be strategically adjusted to affect business growth and other positive outcomes.

Sound like a better way to do business? Here’s an overview of how to get there.

Mature Companies Use Unit Costs As Levers That Control Outcomes

First, let’s talk about why we should get rid of the notion that cloud costs are a necessary evil to be minimized.

In evolved companies, rising costs aren’t necessarily bad. In fact, they’re often a direct result of increased product usage and the corresponding revenue growth that accompanies that usage. Mature companies have the FinOps know-how to evaluate spikes and determine when the increases are proportional to revenue growth or when spending should be dialed back.

But how do mature companies evaluate costs to know whether they’re healthy or unhealthy expenditures? It all comes down to unit economics.

Mature companies have the FinOps know-how to evaluate spikes and determine when the increases are proportional to revenue growth or when spending should be dialed back.

To know whether your cost increases are due to good things (or inefficiencies you should fix), you’ll need to break cloud costs down into individual metrics such as costs per customer, per feature, per workload, per region, or anything else that makes sense in your business context.

Ideally, track as many metrics as possible — including on the revenue side. You want to know how much money each customer, feature, product, or service brings in.

The direct result of understanding your unit economics is that you’ll gain a clearer picture of your company’s overall spending efficiency in comparison to revenue.

That alone is incredibly useful. But it also has some important side effects that, altogether, enable your organization to make this critical shift in perception:

Shift toward focusing on unit economics, and your conversations will pivot from, “Why are our cloud bills so high?” to “How efficiently are we delivering value per dollar spent for each customer or transaction?”

Related: Stop Asking What AI Costs, Ask If It Is Worth It

Suddenly, you’re able to switch from blindly trying to reduce costs to strategically optimizing business outcomes.

Unit economics gets your whole organization on the same page

I often remind clients that cost by itself lacks all business context.

Operating only in terms of total cost can become a friction point in any organization because different parts of the business view costs differently.

The Finance team might balk at a line item for millions of dollars on the AWS bill but without knowing what that spend relates to — how many customers, how many transactions, how much product usage, etc. — it’s impossible to make informed business decisions. They won’t know if the higher cost is due to supporting hundreds of new customers, or if you’re simply overspending.

The ultimate goal of FinOps is to remove or lessen these friction points.

You can do this by operating in terms of unit cost, not just cost as a big umbrella.

Practically, this means setting realistic budgets, actively identifying anomalies, and closely watching trends all through the lens of unit costs.

Start your unit economics journey here

Getting to a mature level takes persistence. Use the below roadmap to get started.

1. Gain cost visibility

Before you can make any real strides, you’ll first need to get a decent picture of your costs. This means right off the bat, you’ll have to confront one of the biggest FinOps barriers in most organizations: tagging and cost allocation.

I would love to tell you that tagging isn’t important. But it is.

Tagging is one of the two core ways to apply business context within cloud environments (the other being account organization). Without robust tagging, you lack the granularity to understand who or what is driving specific costs.

Engineers by themselves don’t typically apply all necessary tags — you need a culture, a policy, and a governance framework that all emphasize adherence to a central tagging strategy.

Your ultimate goal is to improve native tagging by engineers, because this is the most effective way to gain visibility and uncover unit costs. However, tools like CloudZero can help you compensate for inconsistent (or nonexistent) tagging, so don’t get discouraged if your company’s tags leave something to be desired.

Another significant tagging hurdle is shared infrastructure. We often see big blobs of shared costs that need to be broken apart before they make sense. This part of the journey involves identifying shared services (Kubernetes, for example), obtaining usage metrics about them, and then attributing costs to the individual users or teams responsible for them.

2. Uncover just one unit metric

Once you have a clearer picture into your costs, choose one unit metric to track. It should be something that makes sense for your business, but don’t get hung up on perfection.

Just pick something and start from there.

I often advise clients to look at cost per API call, per tenant, or per batch job.

This initial metric will serve as the first companywide unit cost, but it’s just the first of many. Share this metric with everyone in the organization and open the door for some productive cost conversations — because you’re bound to get some feedback.

3. Dive deeper

When you start showing people unit costs, you’ll inevitably encounter a concept called correlation.

Is the chosen unit metric actually correlated with its associated costs and the teams responsible?

If you hold teams accountable for their spend, you’ll start to hear correlation pushback fairly quickly. People will naturally tell you, “Well, this isn’t a great unit metric for me. It would be better if we could track X instead of Y.”

This is where Cloud Cost Center of Excellence teams come into play. As the designated FinOps liaisons in your company, the CCOE team can work with different stakeholders to find demand drivers and usage metrics that closely align with their functions.

Keep having conversations with your staff, and you’ll soon have a whole list of relevant unit metrics your teams want to know more about.

4. Collect more data

The more data you can access, the better. The good news is you can derive usage and cost metrics from the data you already collect. CloudZero offers a nearly effortless way to collect and organize cost data.

But you probably also have access to an untapped goldmine of platforms that can help you dive more deeply into unit costs:

This data is what will enable you to actually track the unit metrics your teams are interested in.

5. Evolve and tailor your unit metrics over time

As your organization matures, you’ll expand from that first global metric you introduced to the whole company into sharing multiple correlated metrics for individual teams or contributors.

If a person or team pushes back on a metric, this is your signal to look for a more relevant, better correlated alternative. But this doesn’t mean you have to abandon the other metrics you come up with.

In an advanced stage, you might track a global company metric and still allow individual team members or groups to work with Finance to identify more specific metrics in addition to the ones you set.

6. Eventually, separate fixed versus variable infrastructure costs

Fixed infrastructure costs are generally not correlated to usage metrics, whereas variable costs should be.

In mature companies, teams are held accountable on two different levels:

The latter ensures that fixed spend remains at an acceptable level and doesn’t grow disproportionately.

Distinguishing between fixed and variable spend isn’t necessarily straightforward if you’re simply looking at your AWS bill. You’ll typically need to combine AWS service data with a map of microservices, applications, and features before you can classify expenses as fixed or variable.

This does demand some rigor with your tagging, as well as a mapping method that connects your application components to their corresponding costs.

Embrace iterative improvements

At any given point, it’s unlikely your whole organization will be at the same point in the FinOps journey. Especially in larger organizations, individual teams will be at different levels of maturity — and some will take steps back while others take steps forward.

You might encounter one team that’s incredibly mature due to focused leadership and well-established practices and another team that’s just starting to implement FinOps principles.

If you have a team that’s excelling, use them as a gold standard within your company. This provides other teams with an example to follow as they develop their own best practices.

A final word of advice: Never let the pursuit of perfection deter you from just taking the next step forward in your journey.

This process will be complex, detailed, and iterative — but it’s also completely manageable and well within your reach. Simply make sure you have the right tools for cost visibility and a strong blueprint for turning that data into actionable unit metrics, and you’ll make progress faster than you think.

Make Uncovering Your Unit Costs Easy With CloudZero

Sure, you can strive for tagging perfection and calculate your unit metrics manually. Or you can have the CloudZero platform fill in the blanks for you with the data you already have in your system.

to see how effortless tracking unit costs can be.

Author: Matt Manger

Matt Manger is CloudZero’s co-founder and VP of Customer Success. With more than 20 years of software engineering experience, Matt leads CloudZero’s FinOps Account Managers (FAM) in making sustainable innovation a reality for every cloud-driven organization.