Australian tech leaders overestimate developer productivity

Australian technology leaders overestimate developer productivity by nearly half, according to research commissioned by Avocado and Dynatrace. The study drew on responses from 154 technology leaders and engineering practitioners at Australian organisations with more than 500 employees.

The findings point to a consistent divide between how executives assess engineering work and how developers describe their daily routines. Leaders were found to overstate the time developers spend building and releasing features, while understating the time spent navigating internal systems, switching between tools, finding information and meeting compliance requirements.

The mismatch also appeared in views on access to production data during development. Some 45% of technology leaders said production telemetry was fully integrated into development tools, compared with 22% of developers. That suggests leaders were more than twice as likely to believe the information was readily available.

Developers also reported losing significant time to tool-switching, yet few leaders said they had a dependable way to measure that loss. The research suggests engineering capacity may be judged on assumptions that do not match the experience of software teams.

Gerardo Barranquero, Chief Executive Officer of Avocado, said the issue had direct business consequences for employers competing for skilled engineers.

"For Australian enterprises operating in a tight and expensive tech talent market, this gap has a real dollar cost attached. Leaders are often making resourcing and hiring decisions based on a picture of engineering capacity that developers don't recognise, which means productivity initiatives, and the budget behind them, can end up aimed at the wrong problem," Barranquero said.

Retention pressure

The study also linked developer experience with staff turnover. It found 74% of technology leaders said developer experience and tooling contributed to engineering attrition somewhere in their organisation.

That matters in a local market where engineering talent remains costly to replace. The data suggests some organisations may be treating attrition as a pay or labour market issue when internal systems and workflows are also playing a role.

"Engineering talent in Australia is hard-won and expensive to replace, and attrition tied to tooling and friction is one of the most avoidable costs in a technology budget. Unlike a competitor poaching on salary, this is a risk organisations largely create themselves, and can largely fix themselves too," Barranquero said.

Platform adoption

Platform engineering has been widely adopted among larger organisations in the survey. The research found 77% of respondents said their organisations were building platform engineering functions or had already established them.

Among those with established programs, 72% reported faster software delivery cycles, 71% improved security outcomes and 66% better visibility across systems. Even so, the study found persistent operational barriers. Some 44% of organisations cited fragmented tooling and systems as a hurdle, while 40% pointed to insufficient observability or production telemetry.

Leaders and developers also differed on whether these efforts were delivering results. While 45% of technology leaders said platform engineering was established and producing measurable value, only 33% of developers agreed.

The results indicate organisations may be meeting internal rollout targets without changing daily work for engineering teams. The study argued that the return on platform investment depends on whether developers can solve problems with fewer tools and less friction, rather than whether adoption milestones have been formally recorded.

"We regularly see platform engineering initiatives that only succeed on paper. They're adopted, standardised and reported up the chain, but they fail to change a developer's actual day. If the platform hasn't reduced the number of tools someone has to open to fix one problem, the investment hasn't landed yet, regardless of what the adoption dashboard reports," Barranquero said.

Local picture

The study was conducted by Ecosystm and was described as being built entirely on Australian data rather than broader regional samples. That distinction matters because much of the research used by technology executives in Australia is drawn from larger international markets, which may not reflect local engineering conditions, hiring pressures or organisational structures.

Darian Bird, Principal Advisor at Ecosystm and author of the report, said the data showed a gap between how leaders still measure software teams and how developers experience their work.

"AI is exposing a truth that engineering teams have long known but rarely prioritised: understanding how applications behave in production matters as much as writing the code itself. Our data shows Australian leaders are still measuring engineering success the old way, by output, while developers are experiencing a very different reality underneath," Bird said.

Barranquero said local data should shape decisions on tooling, platforms and AI-related investment.

"Most of the developer experience data available to Australian leaders is drawn from much bigger, very different markets, so it doesn't always reflect what's actually happening inside a local engineering team. That's exactly the gap we wanted this data to close. Australian leaders should be making platform, tooling and AI investment decisions based on what their own teams are experiencing, not an average pulled from markets many times our size," Barranquero said.