Agile Effort Estimation - How do human and organizational factors shape agile effort estimation?
Agile effort estimation is more than a technical task, it's shaped by team dynamics, communication, and decision-making. This thesis explores how factors like psychological safety, seniority, and rituals such as Planning Poker influence estimation accuracy and consistency.
Topic
Agile effort estimation is often treated as a purely technical task, yet estimation accuracy is shaped by how teams interact, communicate, and make decisions. This Master Thesis investigates how human and organizational factors, such as psychological safety, seniority, and estimation rituals like Planning Poker, impact the accuracy and consistency of agile effort estimation.
Relevance
Accurate effort estimation is critical for agile planning, team capacity forecasting, and sustainable delivery. However, many teams struggle with biases, uneven participation, and unspoken assumptions. This research reveals how estimation is not just a numbers game, but a socio-technical process shaped by trust, facilitation, and inclusion, offering agile practitioners new perspectives to improve delivery predictability.
Results
The study found that teams with high psychological safety and inclusive rituals such as Planning Poker produced more consistent estimates. In contrast, teams where senior voices dominated or where estimation was informal showed signs of anchoring and groupthink. Non-estimation benefits, like knowledge sharing and team alignment, were just as impactful for improving estimation reliability.
Implications for Practitioners
- Encourage structured estimation rituals (e.g., Planning Poker) to ensure balanced participation.
- Reinforce psychological safety through no-blame culture and open dialogue.
- Regularly review and adapt estimation practices during retrospectives.
- Use reference stories to create shared understanding—but update them periodically.
- Be aware of seniority bias and design estimation formats that prevent anchoring.
Methods
The study followed a qualitative multi-method design. Data was collected from five agile teams via ten semi-structured interviews (with Scrum Masters and Developers), direct observations of refinement sessions, and analysis of sprint velocity data. Thematic analysis was used to identify recurring patterns across teams. This triangulated approach ensured rich insights into how team culture, facilitation, and interpersonal dynamics shape estimation behaviour in real-world agile settings.