Application performance always matters. Sometimes critically, sometimes subtly, but it’s never irrelevant. Developers are right to warn against premature optimisation, but ignoring performance until it becomes a real problem can be just as costly.

Performance engineering is not just about speeding up slow code. It is about building habits and processes that keep performance in mind throughout the development lifecycle, rather than treating it as something to fix under pressure. Many teams understand the ideas, but struggle to apply them consistently when performance really matters.

In this talk, we will look at how to make performance engineering a sustainable part of everyday development. We will cover practical approaches for making performance visible and actionable, including production monitoring, performance budgets and SLOs, early feedback loops, and shared ownership across the team.

We will then walk through a single real-world .NET optimisation example. Using monitoring, profiling, and benchmarking data, you will see how an actual performance problem is identified, investigated, and improved, and how to decide whether further optimisation effort is justified. Tools such as dotTrace, dotMemory, and BenchmarkDotNet will be used to validate the results.

We will also discuss the trade offs teams face in practice. This includes balancing performance work with feature delivery, deciding when “good enough” is good enough, making the case for optimisation to stakeholders, and improving efficiency without sacrificing code quality.

By the end of the session, you will leave with practical .NET performance techniques and a clear framework for knowing when optimisation is worth the effort, and when it is not.