Cloud-enabled: what digital sustainability demands (Guest blog from GoCodeGreen)

This blog was written by Eric Zie, Founder and CEO, GoCodeGreen

Cloud computing is widely viewed as a cornerstone of digital transformation—promising scalability, efficiency, and resilience. Recently, it has also been presented as a pathway to sustainability. Cloud providers promote cleaner grids, energy-optimised infrastructure, and reduced operational overheads. And while these benefits can be real, the latest empirical data shows they are not enough when undertaking a deep decarbonisation of digital services.

If sustainability is the goal, then the cloud must be understood not as the solution but as an enabler. Real emissions reductions are only delivered when cloud migration is paired with software modernisation, architectural redesign, and operational maturity. Without these, organisations risk simply shifting inefficiencies from one environment to another—missing the full opportunity of cloud-native transformation.

This conclusion is based on a dataset of over 200 lifecycle assessments of digital platforms, conducted independently using a standards-aligned method by GoCodeGreen. These assessments spanned cloud and on-premise deployments, across sectors and system types, and captured the emissions footprint of digital products across three key lifecycle phases: design, release, and use.

Cloud reduces emissions—but only when used intentionally

If sustainability is the goal, then the cloud must be understood not as the solution but as an enabler.

The comparative data confirms that cloud hosting does reduce emissions versus on-premise infrastructure—by up to 14% in design-related impact, 16% in release, and 10% in operational use when best-practice patterns are adopted. However, the same dataset also reveals that the gap between cloud platforms and between cloud users is larger than the gap between cloud and on-prem.

This matters. It shows that simply moving to the cloud is no guarantee of sustainability. Two platforms using the same hyperscaler—even the same region or service—can deliver radically different lifecycle performance depending on how they are designed, built, and managed. Cloud efficiency, in other words, is not delivered by the provider. They enable it, and the engineering teams that build on top of them realise it.

Why architecture and engineering define cloud impact

The data shows that the most efficient platforms—in terms of emissions per lifecycle stage—were those that fully embraced cloud-native patterns. They did not just “lift and shift” workloads. They rebuilt them around architectural discipline, right-sized compute, and intelligent runtime scaling.

These platforms used event-driven services and serverless functions to align usage with demand. They deployed infrastructure-as-code to reduce environment drift and teardown waste. And they embedded sustainability into design constraints, taking advantage of regions with cleaner grids and services with lower overhead.

In contrast, platforms that retained monolithic design, relied on persistent infrastructure, or failed to optimise release cycles saw minimal improvement—and in some cases, even increased impact in release and use phases due to inefficient development and support practices.

Use-phase efficiency depends on upstream decisions

One of the most consistent findings in the data was the importance of upstream design and delivery decisions in shaping downstream emissions. In both cloud and software assessments, the Use phase (post-deployment support and operations) accounted for over 30% of total lifecycle impact. But that impact was not fixed.

Teams that embedded observability discipline, tuned telemetry, and aligned monitoring with real usage patterns consistently reduced use-phase emissions by up to 6%. But these operational gains were only achieved when supported by clean architecture, agile pipelines, and low-latency feedback loops—all of which depend on how the system was designed and built in the first place.

This study reinforces a central insight: the efficiency of a cloud system at runtime is determined long before it is deployed. The cloud can only execute what the engineering team creates. If that creation is inefficient, elastic infrastructure cannot undo it.

Cloud as a sustainability multiplier, not a substitute

Rather than viewing the cloud as the solution to digital emissions, the evidence points to a more accurate framing: the cloud is a multiplier. It accelerates and amplifies the impact of good engineering—but also the consequences of poor design.

Well-architected platforms that use automation, modularity, and runtime scaling gain more from cloud migration than those that do not. The same is true of teams that manage deployment velocity and environment hygiene using continuous integration and delivery (CI/CD) pipelines. For these platforms, cloud becomes a sustainability enabler—delivering faster releases, lower idle time, and measurable carbon reduction.

However, organisations that view cloud migration solely as a destination may experience limited benefits. This is why the narrative must shift. Cloud computing is a means to sustainable digital systems—not an end in itself.

The role of measurement and lifecycle visibility

Another key learning from the 200+ platform dataset is the need for full lifecycle visibility. Impact is not evenly distributed. In most systems assessed, emissions from the design and release phases exceeded those from use. This means that improvements in software development—such as automated environment teardown, reduced test redundancy, and streamlined release management—should not be treated as marginal; they are central to reducing total footprint.

Moreover, impact reduction is cumulative. The data models show that systems with integrated sustainable practices across architecture, build, and operations achieved lifecycle reductions of up to 53%. In contrast, those who focused only on runtime optimisation—even with clean infrastructure—rarely exceeded 20%.

Therefore, organisations must treat sustainability as a lifecycle discipline. It must begin at design, extend through delivery, and persist in operations. Cloud migration opens the door. But only integrated engineering walks through it.

What this means for cloud strategy

As regulations tighten and expectations rise, organisations must go beyond migration. They must modernise. They must measure. And they must integrate sustainability into every stage of the software lifecycle.

For IT leaders and cloud strategists, the implications are clear:

Cloud migration should not be considered synonymous with decarbonisation. Hosting in the cloud enables emissions reduction, but only if paired with software modernisation and architectural redesign
Measurement must span the full software lifecycle — from design and development through to live operations — not just runtime telemetry. Scope 3 reporting and sustainability claims require insight into design, build, and use—including development activity and support overhead.
GreenOps begins with engineering, not dashboards. Telemetry supports optimisation — but cannot recover emissions from poor design choices. Dashboards and telemetry help manage operations but cannot compensate for poor design. Efficiency begins in the codebase, not at the monitoring console.
Sustainable software is not slower software. The highest-performing platforms delivered faster releases, lower support costs, and fewer outages—evidencing that sustainability and velocity are aligned.

Cloud computing will remain essential to digital sustainability. But treating it as a carbon cure-all is misleading. It is the software that runs on the cloud—and the decisions that shape that software—that determine real-world environmental impact.

As regulations tighten and expectations rise, organisations must go beyond migration. They must modernise. They must measure. And they must integrate sustainability into every stage of the software lifecycle. The evidence is clear: only cloud-native maturity — not mere cloud presence — delivers meaningful, measurable emissions reduction.

Cloud Week 2025

Check out more insights on a range of key topics related to Cloud

Find out more

Call for contributions: A sustainable future for Cloud, Data and AI

We are asking techUK members to submit case studies and success stories highlighting the latest innovation and best practice that supports a sustainable approach to cloud, data and AI.

Find out more