Estimates – a necessary evil? - Erik Thorsell

The post argues that product owners need estimates to prioritize backlogs, compare large and small features, and coordinate releases across teams. It also says estimates are useful for weighing technical debt against new work and for understanding how long feature implementation and integration may take.

It then argues that developers often dislike estimates because uncertainty, hidden dependencies, and technical debt make precise forecasting difficult. The main conflict comes when estimates are treated as deadlines or predictions, which pressures developers and reduces psychological safety. The suggested response is to keep sharing updated estimates as work progresses.

Reading notes

#

• Product owners use estimates to prioritize backlog items because demand usually exceeds delivery capacity.
• Estimates help decide whether to include a large feature or several smaller ones before a communicated release date.
• In larger products, multiple product owners need estimates to coordinate release content and timing across main and team backlogs.
• Technical debt affects how long features take and why developers may want to improve code while working on it.
• Developers dislike estimates because they are asked to foresee uncertainty, hidden dependencies, and integration problems.
• The post says estimates are often reused as deadlines, which makes teams pad their numbers and become less willing to estimate.
• The core problem is expectations created by advance communication of features and dates.
• The post recommends continuous updates to estimates during development instead of treating the first estimate as fixed.
• It says DevOps flow can help reduce delivery risk and surface issues in the product lifecycle.