The Anatomy of Gross Margin in SaaS

The SaaS model is very different from traditional software businesses. It assumes that applications run in a software provider's environment. This means that infrastructure and operating costs shift from end-customer to the SaaS provider's side. Which is why it's not surprising that software providers are always looking to optimise gross margin as much as possible.

In a nutshell, gross margin is how much revenue you keep after delivering your service and it is calculated as (Revenue - Cost) / Revenue * 100. As you can see from this formula, if cost of running your SaaS application grows proportionally with your revenue, your gross margin remains the same. So, if you want to improve your gross margin, you need your revenue to grow faster than your cost. In this article I’m going to explore what is included in SaaS cost and how application architecture influences it.

Infrastructure costs are the most obvious part of your SaaS spend. This typically includes compute resources, storage and third-party services. Therefore, these are the most obvious targets for SaaS cost optimisation. One of the most popular operations frameworks, FinOps , defines two types of infrastructure cost optimisations: usage optimisation and rate optimisation.

Usage optimisation refers to using as few resources as possible for delivering your service. Even though consumption itself is not a bad thing, at the end of the day, when your business grows, your resource consumption is also usually growing. However, you need to make sure that you continuously review your resource usage volumes to eliminate waste.

The main sources of waste are inefficient software and underutilised resources. Review your application consumption regularly to make sure it uses resources that it really needs. You may also be surprised by how often software providers forget to shut down resources when they don't need them anymore. Monitor your resource utilisation to identify potential savings. Services like AWS Compute Optimiser  and Amazon S3 Storage Lens  can help you with this. Shut down idle resources and rightsize underutilised ones to make sure you pay for only what you use.

Rate optimisation helps you pay less for resources that you already use. Many service providers offer significant discounts for long term commitments. Savings Plans  and Reserved Instances  are examples of such commitment-based discount programmes.

Model ROI before making any commitments and compare it with on-demand pricing with autoscaling. Track your commitment coverage and utilisation metrics, and don’t forget to review your commitments regularly. Ideally, you need to review your commitments based on metrics instead of using a calendar-based cadence.

Upfront payments can offer even bigger discount opportunities. The more you pay upfront, the bigger the discount you receive.

Sometimes service providers let you use spare capacity at substantial discounts, up to 90% cheaper than on-demand prices. An example of this offering is Amazon EC2 Spot Instances . They're ideal for fault-tolerant workloads, such as batch processing and background jobs, where occasional interruptions can be handled without impacting users.

And finally, if your resource consumption is significant, you can leverage private pricing agreements with your service providers.

Operating costs are quite often underestimated among software providers. The cost of human resources in many cases comparable to infrastructure cost and deserves special attention when optimising.

SaaS operations include all the activities that keep you application up and running, such as tenant lifecycle management, configuration, troubleshooting and analytics. The two keys for efficient SaaS operations are tools and automations.

Tools are the backbone of efficient SaaS operations. The right tools simplify processes across infrastructure, tenant configuration, and troubleshooting. Monitoring, billing, analytics, and support tools make day-to-day operations manageable even when the number of customers grows.

Instead of engineers and operations chasing data across multiple systems, tools provide a shared source of truth and ready-to-use workflows. This shortens workflow execution times and reduces operational overhead. Over time, a well-chosen toolset enables SaaS companies to grow more smoothly, keeping operating costs under control while maintaining a high-quality customer experience.

In SaaS, these tools usually live under the control plane. A strong control plane centralises configuration, orchestration, and visibility, making day-to-day operations smoother and more predictable. Instead of scattering tools across services, operations team gets a single place to manage changes safely and consistently. That's why investing early in a solid control plane pays off later.

As the number of customers grows, manual interventions simply don't scale effectively, which is why automations are significant levers for improving operational efficiency in SaaS. They remove repetitive manual work from processes like customer onboarding, billing, customer support and incident response. This speeds up execution and lets small teams operate applications that would otherwise require far more people.

Customer lifecycle automations cover everything from onboarding and offboarding to tier level changes. Instead of manually creating accounts, provisioning resources, or cleaning things when users leave, these workflows run automatically and consistently, giving your operations team time back.

Customer support automations also take the repetitive work off your operating team by handling predictable stuff automatically. Tickets are auto triaged, and diagnostics like logs and recent user activity are attached to tickets upfront, while self-service workflows let customer fix common issues on their own. In the result your operations team spends time on exceptions and edge cases, not repeating the same routine tasks all day.

Incident management automations are about letting machines do the boring, time-critical prep work before humans step in. Post-incident investigations, such as aggregating relevant data and turning it into a first draft of an investigation report, can be done by machines. The end goal is to let automations handle preparations, so humans can focus on judgement and decision-making.

An effective development cycle helps protect SaaS gross margin by keeping costs predictable as your business scales. When engineering teams spend more time building customer value and less time fixing issues, it allows your revenue to grow faster than engineering costs. There are some practices that help can you improve your development cycle in SaaS.

Avoid per-customer customisations. They can feel like a quick win, but they tax your application code over time. Each bespoke change adds complexity and makes the codebase harder to test, evolve and release. Sometimes to the point where engineering team is effectively supporting multiple products. By designing configurable, shared features instead, you ship once for everyone and keep your development cycle efficient as your business scales.

Avoid version pinning. Keeping everyone on the same SaaS version simplifies debugging, testing, and shipping changes. When customers run different versions, even small changes turn into complex and risky releases with special-case handling. A continuously updated SaaS lets engineering team focus on building new value instead of maintaining old versions.

Don't reinvent the wheel. Managed services and off-the-shelf solutions are cheaper to maintain and usually more reliable than custom alternatives. You save engineering time upfront and avoid long-term costs like bug fixing, security patching, and on-call support. That means fewer engineers are needed to support the same revenue, faster time to market, and more predictable costs.

In a SaaS model, where you run customers' workloads in your own environment, gross margin significantly contributes to profitability. And as you can see from this article, there are multiple ways to improve your cost efficiency. So, it's not surprising that the best results can be achieved by combining different cost optimisation practices. Start from the lowest-hanging fruits and move gradually towards more complex optimisations until they have a positive return on your investments.