Australian developers are losing half their day, most leaders have no idea
Mon, 13th Jul 2026 (Today)
Australian organisations are investing in platform engineering, cloud modernisation, and AI-assisted development at pace. But beneath these investments is an invisible drag eroding returns: developers are spending the majority of their working day on everything except building software.
Ask any engineering leader how productive their developers are, and most will offer a confident answer. Ask the developers themselves, and you will get a very different one. According to IDC, developers spend just 16 per cent of their time on core development work - the rest is consumed by CI/CD processes, security, and performance and infrastructure monitoring. Atlassian's 2025 Developer Experience Report, surveying 3,500 developers and managers, arrived at the same figure independently. At enterprise salary levels, that lost time represents a substantial hidden cost per developer each year - one most boards have never seen framed that way.
Is this playing out in your organisation? Ecosystm is running an independent study -commissioned by Avocado Consulting - on developer experience in Australia, capturing perspectives from both engineering leaders and the developers they manage as AI adoption accelerates. Take part / Share your perspective now
The gap between what engineering leaders perceive and what developers actually experience is one of the most commercially significant problems in Australian technology organisations today.
The platform engineering promise and its blind spot
The intent of investing in platform engineering is sound: standardise the developer experience, reduce cognitive load, accelerate delivery. But a critical capability gap is quietly undermining those investments before they are fully realised - and it starts with observability.
Without genuine visibility into how software behaves in production – the root cause of degradation, its blast radius, ownership, and business impact – developers spend their time in incident management bridges rather than innovating. Legacy systems with poor instrumentation compound the problem, creating blind spots where issues accumulate silently and unresolved incidents resurface later at greater cost. Platform engineering initiatives that ignore observability risk creating modern delivery pipelines without the operational insight needed to run them safely.
AI in production: when AI moves faster than your ability to see it
Australian developers are adopting AI coding tools and integrating large language model components into production systems faster than most organisations can govern them. The productivity gains are real, but so is the risk: AI components can fail silently, behave unpredictably under load, and generate costs that are difficult to attribute or forecast. Many organisations lack visibility into how those components perform, what resources they consume, or whether they behave as designed.
AI can accelerate productivity, but it also introduces operational complexity that leaders must be able to detect, understand, and manage. That complexity extends beyond reliability and cost. When AI-native systems are not well understood, organisations increase the risk of inadvertently exposing sensitive information, missing guardrail breaches, or carrying production behaviours they cannot adequately govern. Without visibility into how these components perform in production, a new category of risk sits inside the development lifecycle.
For engineering leaders already operating with an incomplete picture of their developers' experience, and under pressure to deliver more with fewer resources, the AI layer adds risk – but also creates significant potential for safe innovation when understood and governed effectively. Observability gives leaders the assurance that AI systems are operating as intended, with the visibility needed to manage risk, maintain compliance, and give developers the confidence to innovate.
Closing the gap requires more than tooling
Organisations navigating this effectively share a common characteristic: they connect observability data to developer workflows, rather than treating it as a separate operational discipline. Platforms such as Dynatrace help connect production behaviour to those workflows in real time, giving teams signals to identify, attribute, and resolve issues earlier. Even when issues inevitably slip through cracks, high-fidelity production data reduces the time and effort spent troubleshooting or reproducing issues, and gives the right teams precise, actionable context. When developers can see critical signals – including cost impact, performance degradation, and AI component behaviour – they make better decisions earlier, preventing many problems from ever reaching production. Productive engineering depends on feedback loops that operate at the speed of development, not the speed of a post-incident review.
For engineering leaders, the first step is not procuring new tooling. It is closing the perception gap: understanding how developers actually spend their time, which processes add toil, where friction accumulates, and where visibility gaps reside. The data can be confronting, but it is also the most useful input a leader has to make meaningful change.
The business case is already there
Developer experience is no longer only a talent retention conversation, though it remains that too. It is also a financial one. The cost of invisible friction is often absorbed unknowingly into the operating model: troubleshooting without adequate signals, navigating tool sprawl, managing AI components without observability, and accumulating technical debt in systems nobody can see clearly. It compounds quietly until it surfaces as delayed delivery, unexpected cloud expenditure, or engineering attrition – outcomes that may not immediately be attributed to a lack of visibility. Australian organisations that treat observability as the connective tissue of their development lifecycle, rather than an operational afterthought, will give developers back more of the day for the work they were hired to do. The productivity dividend is significant, measurable, and long overdue.
Want to help shape this conversation? Ecosystm's Developer Experience in Australia Study (commissioned by Avocado) is gathering views from C-suite and technology leaders alongside the developers on the ground - exploring the perception gap, observability maturity, and AI's operational risks and rewards. Contribute your insights / Access the study: https://ecosystm.questionpro.