SaaS companies decarbonisation guide

Software as a Service (SaaS) companies are increasingly recognising the environmental impact of their digital operations. The carbon footprint of cloud-based services extends beyond just server energy consumption, it encompasses the entire software lifecycle from development to deployment, and from ongoing operations to end-user experience. With the ICT sector responsible for between 1.5% and 4% of global carbon emissions according to the World Bank, SaaS providers face both a responsibility and an opportunity to lead in sustainable digital transformation.

This guide outlines practical strategies that SaaS companies can implement to reduce their carbon footprint while maintaining, and even enhancing, performance, reliability, and user experience. By focusing on sustainable software engineering practices, we'll explore how optimising your code, infrastructure, and development processes can yield significant environmental benefits alongside operational efficiencies.

Understanding emissions in SaaS operations

The carbon footprint of SaaS operations primarily stems from data centre energy consumption and the efficiency of software code. These emissions typically fall into three categories:

Scope	Emission source	Relevance
Scope 1	Direct emissions	Usually minimal for pure SaaS companies; may include on-site generators or cooling systems using refrigerants
Scope 2	Purchased electricity	Significant for companies with owned/leased data center space; includes power for servers, cooling, and office operations
Scope 3	Value chain emissions	Usually the largest portion; includes cloud services purchased from providers, employee commuting, and emissions from customer usage of software

For most SaaS companies, the largest portion of emissions comes from Scope 3 sources, particularly from purchased cloud services. According to analysis published in Nature Communications by Klaaßen & Stoll (2021), emissions from IT software companies are often underreported due to inconsistencies in measurement methodologies and boundary incompleteness.

When a company moves from on-premise infrastructure to cloud services, emissions that were previously classified as Scope 1 or 2 shift to Scope 3, as noted by Mytton (2020). This accounting shift doesn't reduce emissions, it redistributes responsibility, which makes understanding your cloud provider's environmental practices critically important.

Emission shift when moving to cloud services

Optimising infrastructure for energy efficiency

Optimising your infrastructure is one of the most effective ways to reduce your carbon footprint. Here are key strategies focusing on cloud resources management:

Implement rightsizing and autoscaling

Server optimisation begins with appropriate sizing and dynamic scaling. Rightsizing ensures you're not provisioning more computing resources than needed, while autoscaling automatically adjusts resources based on demand.

According to research by Wang et al. (2023), implementing full scaling automation in data centers helped reduce 947 tons of CO₂ emissions and save over 1.5 million kWh of electricity during a high-demand shopping festival. For SaaS companies, this means:

Regularly audit your virtual machines and containers to identify over-provisioned resources
Configure autoscaling policies based on actual application metrics rather than time-based rules
Consider using "spot instances" for non-critical workloads, which utilise unused cloud capacity at lower rates

Server utilisation and energy consumption relationship

Adopt containerisation for greater efficiency

Containerisation offers significant efficiency advantages over traditional virtualisation. Technologies like Docker and Kubernetes allow for better resource utilisation and reduced energy consumption:

Containers require fewer resources than full virtual machines, allowing more applications to run on the same hardware
Orchestration tools optimise container placement for maximum server utilisation
The lightweight nature of containers enables faster scaling up and down in response to demand fluctuations

A real-world example comes from Nordstrom, as cited by Férdeline (2021). The retailer transitioned from thousands of virtual machines with just 4% average utilisation to using Kubernetes, which increased utilisation to 40% while reducing their infrastructure footprint by 90%, equating to substantial carbon emission reductions.

Select sustainable cloud regions

Not all data centers are created equal when it comes to carbon intensity. According to Climatiq, the carbon intensity of cloud regions can vary by a factor of 5 or more depending on the local electricity grid mix.

When deploying your SaaS application:

Use tools like Google's Cloud Region Picker to evaluate the carbon intensity of different regions
Balance carbon considerations against latency requirements for your user base
Consider multi-region deployments that prioritise low-carbon regions for non-latency-sensitive workloads

Strategy	Impact	Implementation complexity
Choose renewable-powered regions	Can reduce associated emissions by 50-90% compared to coal-powered regions	Low - requires configuration changes only
Implement container-based architecture	Can improve resource utilization by 5-10x over traditional VMs	Medium - requires architecture and deployment changes
Optimize instance sizes/autoscaling	Typically reduces cloud resources by 20-40%	Medium - requires monitoring and configuration tuning

Writing efficient, sustainable code

While infrastructure optimisation is crucial, the efficiency of your actual software code has a direct impact on resource consumption and energy use. Sustainable software engineering practices can substantially reduce your carbon footprint while improving performance.

Optimise database operations

Database operations often represent a significant portion of energy consumption in SaaS applications:

Implement proper indexing to reduce query execution time and resource usage
Use caching strategies to minimise repeated database queries
Consider NoSQL or purpose-built databases for specific workloads rather than forcing everything into a traditional relational database
Optimise query patterns to reduce data transfer volumes and processing requirements

Employ code profiling and refactoring

Identifying and eliminating inefficient code is essential for sustainability. According to Shi et al. (2024), optimizing computational processes in large language models achieved up to a 157 times reduction in carbon footprint.

For your SaaS application:

Use profiling tools to identify performance bottlenecks and resource-intensive code sections
Apply performance-driven refactoring to resolve common antipatterns
Consider utilising Just-In-Time (JIT) compilation for appropriate languages to reduce energy consumption

Research by Arcelli et al. (2017) demonstrates how refactoring techniques like "Move Operation to New Component and Node" can resolve performance antipatterns that cause bottlenecks in software operations, leading to both improved user experience and reduced energy consumption.

Choose appropriate programming languages

The choice of programming language can significantly impact energy consumption. A study comparing native and web applications found that native applications consistently consumed less energy, as shown by Horn et al..

Consider these factors when selecting programming languages:

Compiled languages (Go, Rust, C/C++) generally produce more energy-efficient executables than interpreted languages (Python, JavaScript)
For interpreted languages, leverage native functions and libraries whenever possible
Balance energy efficiency against development speed and maintainability requirements

The energy impact of language choice becomes increasingly significant at scale. For high-volume SaaS operations, even small efficiency improvements per transaction can translate to substantial energy and cost savings.

Implementing sustainable CI/CD pipelines

Continuous Integration/Continuous Delivery (CI/CD) pipelines represent a significant portion of compute resources in many SaaS development environments. Optimising these pipelines can reduce both development-time emissions and improve developer productivity.

Accelerate test execution

Tests are necessary for quality assurance, but they can consume substantial resources if not optimised:

Run tests in parallel across multiple CPU cores instead of sequentially
Implement test segmentation to run only the tests affected by code changes
Leverage caching mechanisms to avoid redundant setup and teardown operations
Consider containerised testing environments that can be spun up and down quickly

Implement event-driven architecture

An event-driven architecture can significantly reduce unnecessary computational workloads:

Replace polling mechanisms with push notifications and event triggers
Implement asynchronous processing for resource-intensive operations
Design microservices to respond only when needed rather than constantly checking for updates

According to Plan A's research on IT decarbonisation, event-driven architectures avoid unnecessary status update requests among services, reducing both network traffic and computational load.

Educating users and extending software longevity

The impact of your SaaS application extends to how your users interact with it. Education and design choices can significantly affect user-side energy consumption.

Promote energy-efficient usage patterns

Provide guidelines to help users minimise their energy consumption:

Enable and encourage "dark mode" for OLED/AMOLED displays to reduce power usage
Design your application to use fewer resources on mobile devices through responsive design
Implement session timeouts and automatic logout for inactive users to reduce background processing

Design for software longevity

Software that requires frequent replacement or updates contributes to higher overall emissions:

Design modular architectures that can be updated in components rather than requiring full replacements
Implement backward compatibility to extend the useful life of your software
Consider the full lifecycle environmental impact when planning major software changes or migrations

User Practice	Energy impact	User benefit
Enable dark mode where available	Reduces screen energy usage by 30-60% on OLED displays	Reduces eye strain in low-light environments
Close unused tabs and applications	Can reduce device energy consumption by 15-25%	Improves device performance and battery life
Download content for offline use rather than streaming	Reduces network energy usage and can lower overall consumption by 40-75%	Ensures access without connectivity and often provides better performance

Measuring and tracking your improvement

To effectively reduce your carbon footprint, you need to measure your current emissions and track improvements over time. Plan A's carbon management platform provides specialised tools for software companies to measure, report, and reduce their emissions.

Key metrics to track include:

Energy usage per user or transaction
Server utilisation rates
Energy efficiency of your code base over time
Carbon intensity of your cloud providers and regions

Carbon reduction journey for SaaS companies

Decarbonising your SaaS operations requires a multi-faceted approach that spans infrastructure optimisation, code efficiency, development practices, and user education. By implementing the strategies outlined in this guide, you can significantly reduce your carbon footprint while often improving performance and reducing operational costs.

The journey toward sustainable software isn't just an environmental imperative, it's becoming a business necessity as customers, investors, and regulators increasingly prioritize environmental performance. Companies that lead in this space will gain competitive advantages through operational efficiency, regulatory compliance, and market differentiation.

For a deeper technical exploration of sustainable software engineering principles, read Plan A's detailed resource on IT decarbonisation. To understand the broader impact of green IT on your company's emissions, explore how green IT can reduce your company's emissions.

Ready to start your SaaS decarbonisation journey? Contact Plan A's experts to learn how our platform can help you measure, report, and reduce your carbon footprint with science-based methodologies tailored for software companies.

‍