CIOs have spent the past decade chasing scale, faster delivery, bigger stacks, and more automation. But in that sprint to modernise, something critical slipped through the cracks: the efficiency of the code itself.
Bloated libraries, AI-generated shortcuts, and unchecked reuse are flooding production environments. What looks like velocity at the developer level becomes waste at scale, especially in cloud and edge systems where every inefficiency compounds across thousands of containers, clusters, and constrained devices.
The result? Massive compute overhead, spiralling costs, and silent energy use no one is tracking, let alone accountable for.
This might have passed as a low-priority concern a few years ago. But as more organisations move to distributed architectures and deploy at scale, the impact of bloated code is becoming harder to ignore.
Every unnecessary line, every unoptimised library, adds to processing time, power draw, and resource consumption. In Australia alone, projections see data centres growing to account for between 8% and 15% of national energy consumption by 2030, and it’s clear to see how these inefficiencies can multiply. In short: what used to be a developer’s shortcut is now an executive’s liability.
While cloud costs can be obscured by procurement cycles, edge computing doesn’t let you hide. With industries like agriculture, mining, and logistics investing heavily in edge and IoT infrastructure, Australia is uniquely exposed to software inefficiency at the physical edge, where power constraints and network costs compound even minor coding waste.
Every watt counts. Every byte of transfer has a cost. There’s no such thing as “elastic” compute on a remote cattle station, and yet code is still being pushed as if it were running in a data centre.
Global momentum, local inaction
Other regions have started to reckon with this. In the EU, regulators are beginning to include digital energy use within broader operational compliance frameworks, such as with the Digitalising the Energy System plan launched in 2022. The aim is to move software performance from the realm of IT into measurable governance standards.
In Australia, we’re not there yet.
Despite efficiency targets, digital operations remain largely untouched by broader organisational strategies. We still treat software performance like a developer concern, not a boardroom one. There are no benchmarks. No requirements. No accountability for how much compute is used to support applications, let alone what that means for overall energy consumption.
But here’s the truth: Bad code doesn’t just run slow; it burns energy and money.
An IBM report found that inefficient coding practices, unnecessary microservices, and overprovisioned infrastructure are directly linked to higher energy use, operational costs, and system inefficiency. What’s more, some programming languages are able to reduce an application’s energy demands by up to 50%.
In a market increasingly pressured by both regulators and investors to improve operational performance, the software layer can no longer be ignored.
From developer concern to boardroom priority
Efficient coding, at its core, is about writing software that does what it needs to, and nothing more. That means limiting dependency bloat. Prioritising efficient algorithms. Using lightweight architectures. And yes, being selective about what AI-generated code actually goes into production.
This is not just a DevOps issue. For coding efficiency to have real impact, it needs structural and cultural support. Platform teams need to embed performance and energy checks into CI/CD. CTOs need to start treating software waste the same way they treat infrastructure waste: as something to be measured, reported, and reduced. CIOs need to stop assuming that efficiency is someone else’s problem.
Let’s be clear, this isn’t about rewriting everything in C or running everything on a Raspberry Pi. It’s about pragmatism. Lighter containers. Smarter build processes. Fewer redundant services spun up “just in case.” Most organisations already have the tooling to start; what’s missing is executive focus.
That focus must be found soon because here’s what’s coming: as digital operations face closer oversight and performance claims are scrutinised more tightly, software bloat won’t just be a nuisance. It’ll be a compliance issue. A reputational risk. A cost centre you can’t defend.
