Liran Grinberg
Has the price tag for innovation become untenable?
When venture-backed companies were chasing growth metrics, higher cloud bills were shrugged off as unavoidable. But the exuberance of the last few years is fading, and investors are adopting a more somber approach to their portfolio companies’ financials. Usage-based cloud and SaaS services, which have become a major cost center, are coming into the spotlight.
Dev teams need to face the music and start being financially accountable for the infrastructure and services they use. Meanwhile, CFOs and CTOs need to get ready to answer some tough questions at board meetings.
The practice of FinOps offers key ideas and tools that let you understand, design and forecast cloud spend in a way that’s aligned with company goals. By applying FinOps principles, companies have an opportunity to significantly improve their gross margins and chart a path toward profitability, as well as alleviate investor concerns around revenue metrics.
Gross margins are a board-level concern
The cloud and app ecosystems that have developed in the past decade help fuel innovation. They enable developers to build features, design experiments and run tests at breakneck pace, with minimal worries about infrastructure. However, this innovation comes with a hefty price tag and leads to a loss of financial oversight as teams struggle to understand which features or customers are driving up costs.
This has become a pressing issue today. Venture capital firms have grown reluctant to pour more funds into money-incinerating businesses. Gross margins, previously seen as a problem to be solved at some indeterminate point in the future, have become an immediate board-level priority.
What’s particularly troubling for investors is how opaque cloud spend tends to be: A single figure might encompass many internal and customer-facing use cases, making it impossible to coherently justify or optimize.
The shift from cost optimization to FinOps
Chasing growth incessantly is part of the reason cloud spend has ballooned in recent years as companies raced to release features and grab market share at any cost. But another key driver was consumption or usage-based pricing models adopted by cloud service providers.
In these models, buyers don’t commit to a large contract in advance. Instead, they are billed “by the meter” for the resources they use, whether that means compute power, scanning a repository for vulnerabilities, or processing a financial transaction. Usage-based pricing has fuelled the tremendous growth of vendors such as Snowflake, and it can be attractive for smaller companies that want to skip the large upfront costs of annual contracts.
But as cloud services become increasingly embedded in products and development processes, usage-based costs have spun out of control. Auto-scaling makes life easier for developers, but it also means a few careless lines of code can add thousands of dollars to your cloud bill. Tying this spend back to business outcomes is arduous. As organizations adopt multicloud architectures and purchase an unprecedented amount of software from different vendors, this situation is only getting worse.
In the meantime, several types of solutions have popped up.
Cloud cost optimization tools such as Granulate and Zesty reduce infrastructure costs by optimizing workloads. These tools provide short-term relief, but they are tactical rather than strategic, as they don’t provide a way for businesses to become data driven in their approach to cloud usage costs.
That’s where FinOps comes in. Developed in recent years and rapidly growing since 2021, FinOps is a discipline and cultural practice. It’s a set of principles and guidelines that enable companies to make data-driven decisions about cloud spend.
As budgets tighten, FinOps can help company leaders gain more control over their cloud costs and show investors a path to profitability. Let’s look at three core FinOps principles and how they can help companies explain and justify costs to their board:
Know your cloud unit economics
Cloud unit economics is a core concept in FinOps. It is the ability to tie cloud costs and revenues together based around an item that creates value for the business. This could be a user making a purchase in a mobile app, a cybersecurity company onboarding a new customer or an active user on a video conferencing app.
While all of these outcomes generate revenue, they also carry the cost of using cloud resources and supporting app infrastructure. Understanding the gross profit generated by selling one unit (known as the contribution margin) requires companies to calculate the marginal cost generated by all of the services that powered the transaction. This includes the usual cloud infrastructure suspects such as AWS and Snowflake as well as SaaS apps, like Stripe, that charge per transaction.
The concept of cloud unit economics is key to quantifying the cloud’s role in your financial performance, forecasting profitability and building a plan for cost optimization. It also supports other profitability metrics such as contribution margin, customer lifetime value (LTV) and customer acquisition cost (CAC).
Knowing your cloud unit economics is key to building an explainable, transparent model of your cloud costs and is the foundation for many other FinOps processes and metrics.
Understand the full picture of costs per customer and segment
When gross margins matter, companies can no longer get away with spending $2 to acquire a $1 LTV customer. While go-to-market teams are well versed at breaking down CAC and LTV metrics and can usually pinpoint the spend on media, creative and sales resources required to generate a sale, product teams can rarely give detailed answers about the true cost of each customer.
This problem exists across SaaS: Even a mobile app user generates cloud costs. But the situation becomes particularly dire when it comes to complex B2B products that allocate a large amount of cloud resources per customer, such as cybersecurity or analytics tools that store and process their customers’ data:
- Acquisition costs are usually very high and not expected to be paid back within the first year of a customer’s subscription. The ability to predict whether a customer will churn is also much more limited at the start of a contract compared to simpler B2C use cases. This adds a high level of uncertainty when attempting to forecast total contract value (TCV).
- LTV is often calculated as CAC (in sales and marketing spend) minus TCV. But this leaves out the cloud infrastructure used to support the customer’s use case. This is much less predictable for a B2B product compared to a mobile app and can also change throughout the customer’s lifetime based on the amount of data being stored or processed.
- While companies will typically try to transfer cloud costs on to the customer at a premium, different licensing models and the unpredictability of usage-based services could consistently eat away at margins. This leads to the CAC never returning itself at a customer’s average tenure.
This may result in a company flying blind, with its low-margin customers hiding in the larger and undecipherable bill for cloud services. A certain cohort of customers might be easier to acquire through marketing campaigns, but their use of cloud resources makes them ROI-negative over the long term.
Understanding the full cost per customer in real time rather than in a post-hoc analysis can both reduce waste and increase profits. Companies can pivot their go-to-market efforts, focusing product and marketing resources on features or audiences that produce better gross margins. In the long term, this would inform year-over-year gross margins and LTV calculations and help shape the broader strategy.
Create a culture of accountability
You can’t improve what you can’t measure, and you can’t ask engineers to drive down cloud costs when it’s impossible to understand what’s generating these costs. If breaking down cloud unit economics for a certain cohort of customers requires an engineering sprint of its own, it will often take the backseat to more pressing product initiatives.
A culture of FinOps means having ready-made access to cloud cost metrics on the one hand and building a culture of accountability on the other. It means knowing which team, which function and which features are responsible for which part of your cloud bill. Once that knowledge is available, leaders need to ensure that engineers are aware of what each line of code is costing the company and that they take this knowledge into consideration before making a commit.
In this sense, FinOps capabilities are just a stepping stone toward broader cultural change. You should regularly update investors of your efforts in these areas and track how your engineering team is becoming more cost-conscious and metrics-driven over time.
In conclusion
DevOps spend can no longer get a free pass as sales and marketing budgets get slashed. Cloud infrastructure and SaaS applications are weighing down company budgets and are predicted to account for the lion’s share of IT budgets by 2025.
FinOps offers a path toward more sustainable innovation and data-driven decisions in product development. It’s not just a matter of technology — responsible cloud spend requires a cultural shift, which doesn’t happen overnight. Companies that are looking to weather the storm should prioritize cloud financial management sooner rather than later.
Comment