The next era of SaaS efficiency: Where sustainability becomes a core metric

SaaS operators today are deeply familiar with the language of cloud economics. They monitor infrastructure spend closely, optimize for efficiency, and understand how architectural choices influence margins at scale.

What is still far less embedded in day-to-day decision-making, however, is an awareness of the environmental impact tied to those same systems.

That gap is becoming increasingly relevant. As software grows more complex and compute-intensive, the energy required to run it rises in parallel. What used to be a largely invisible externality is now starting to show up as both a sustainability concern and a strategic business consideration.

The shift required is not dramatic, but it is meaningful: alongside cost per workload, teams need to develop an intuition for carbon per workload.

In earlier phases of SaaS, marginal costs were close to negligible. Once a product was built, scaling to additional users required relatively little incremental resource. That dynamic has changed.

Modern applications depend on continuous data processing, persistent background tasks, real-time analytics, and increasingly, AI-driven functionality. These systems consume energy continuously, not just at the moment of user interaction, and as a result, the cost structure of SaaS is no longer purely financial.

Financial and environmental expense

This is where GreenOps starts to take on a more practical meaning. Rather than being treated as a compliance exercise or a brand consideration, it has to become part of how successful companies operate their infrastructure.

Reducing waste in systems, whether that is idle compute, redundant storage, or inefficient workloads, has a dual effect. It lowers costs while also reducing emissions and long-term exposure to regulatory and energy price risks.

In that sense, sustainability is not a separate initiative, but an extension of good operational discipline.

Where I work, I see the same pattern again and again: convenience winning over efficiency in the early stages of growth. Servers are spun up “just in case”, data is endlessly duplicated, and workloads run around the clock whether they create value or not.

These decisions are usually byproducts of moving quickly rather than intentional inefficiencies. But over time, they accumulate into a structural cost base that is both financially and environmentally expensive.

Revisiting decisions

As systems mature, the opportunity lies in revisiting these decisions. Choices around caching strategies, job scheduling, model selection, and data retention policies begin to carry more weight.

They are no longer purely technical optimizations, instead influencing how much energy a product consumes and how efficiently it scales. An AI model that is oversized for its task, for example, does not just increase inference costs, it also drives unnecessary energy usage with every request.

Better visibility is helping to bring these trade-offs into focus. Cloud providers are starting to expose more detailed energy and usage data, while specialized tools estimate emissions across different services and regions.

This does not create perfect measurement, but can enable comparison. When teams can see carbon metrics alongside financial ones, it changes how decisions are made.

AI accelerates this shift even further, because AI-native SaaS does not merely store and serve data - it is constantly computing.

Tasks such as inference, embeddings, classification, and background processing introduce variable operating costs that grow with usage, meaning that more customers no longer just translate into more revenue; they also require more GPUs, more power, and generate more emissions.

Many teams are surprised when they look closely at where their AI energy is actually going, as it is rarely the obvious places. It is background processes, retries, oversized models used for small tasks, and data that no one questions anymore.

This is where FinOps and GreenOps converge most clearly, because designing efficient models, implementing smart batching and caching, and right-sizing infrastructure are as much climate strategies as they are cost strategies.

Efficiency as a constraint

There are plenty of examples of software being designed today with efficiency as a constraint rather than an afterthought. Platforms like DAKboard illustrate this well.

Built to power always-on digital wall calendars, often on low-energy Raspberry Pi hardware, they require careful consideration of how frequently family dashboard data updates, how much processing is necessary, and what can be handled at the edge rather than in the cloud.

They also champion a 'Bring Your Own Device' (BYOD) model. By allowing users to repurpose existing screens and older tablets, DAKboard gives hardware a second life rather than requiring the manufacturing of brand-new displays.

This kind of design discipline naturally limits both infrastructure costs and energy consumption. It highlights that not all SaaS needs to follow a trajectory of ever-increasing compute intensity, and that thoughtful product design can align user experience with operational efficiency from the outset.

ESG has become an important framework for many organizations, but reporting alone does not guarantee efficiency. A company can meet disclosure requirements and still run systems that are fundamentally wasteful.

The more meaningful question is whether sustainability is reflected in how the product itself operates, and whether growth is being achieved in a way that is resource-aware and resilient over time. When acquiring a SaaS company, I’m looking for a balanced approach to business that prioritizes both growth and impact.

Clean energy investment is not accelerating because it feels good, but because it makes economic sense. Today, 55% of low-carbon technologies are already cost-competitive in most situations, or will be soon, and another 10% are only marginally more expensive.

From 2016 to 2024, companies pursuing green growth achieved higher revenue valuations, according to Boston Consulting Group analysis. So sustainability is no longer a trade-off, it is increasingly a competitive advantage.

Climate impact and economic value

I similarly look for technologies where climate impact and economic value are mutually reinforcing. In this space, that means looking beyond the code to the physical components of the digital age.

While software optimization is essential, given the massive energy requirements, we believe a fundamental rethink of the hardware stack is necessary.

This includes investing in next-generation electronics, advanced cooling systems, and novel computing architectures designed to handle AI workloads with a fraction of the traditional power draw. By attacking inefficiency at the silicon and thermal layers, we are investing to decouple AI’s growth from rising energy consumption and emissions.

Sustainability is no longer a “nice to have” when we look at acquisitions and scaling strategies. Efficient infrastructure, disciplined cloud usage, and thoughtful architecture tell us something about management quality. They show whether teams understand leverage, long-term cost, and responsible growth.

Looking ahead, the definition of efficiency in SaaS is expanding. It is no longer just about reducing spend or improving performance in isolation, but understanding how those improvements translate into broader impact.

Companies that treat GreenOps as part of core operations, alongside security, maintainability, performance, and scale, are simply better prepared for the future. Sustainable SaaS is not slower SaaS; it is smarter SaaS.

These companies will be better positioned to navigate future constraints - whether those come from regulation, energy markets, or customer expectations, with more durable, future-ready software businesses.

We've featured the best green web hosting.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit