Description
A structured, non-punitive review of an adverse event that reconstructs what people knew and why their actions made sense, to improve systems rather than assign blame.
Medicine's morbidity and mortality conference has reviewed adverse outcomes for over a century. Software engineering adopted the blameless postmortem, grounded in Sidney Dekker's 'new view' of human error: people's actions made sense to them at the time given their goals, knowledge and focus of attention. A rigorous review reconstructs a timeline, explores local rationality through interviews, identifies contributing factors (including interface and process design), and produces specific action items with owners. For UX, user-facing incidents (a confusing change causing data loss, a dark pattern complaint, a support surge) deserve the same treatment, with users' experiences reconstructed alongside internal decisions.
Attributes
- Discipline
- Human Factors & Error
- Loss type
- Siloed
- Stage
- Post-launch
- Data
- Qualitative
- Effort
- Days
Steps
- Declare a review for any user-impacting incident above an agreed threshold, and set a blameless ground rule.
- Reconstruct a detailed timeline from logs, support tickets, chat records and user reports.
- Interview involved staff and, where possible, affected users, using questions about what they saw, expected and were trying to do.
- Identify contributing factors across design, process, tooling, communication and organisational pressure, avoiding 'human error' as a conclusion.
- Write the review using a standard template and agree action items with owners and dates.
- Track actions to completion and periodically analyse reviews in aggregate for recurring themes.
Sources
- Dekker, The Field Guide to Understanding 'Human Error' (3rd ed., 2014)
- Beyer et al., Site Reliability Engineering (2016), chapter on postmortem culture
- Allspaw, Blameless PostMortems and a Just Culture (2012), Code as Craft blog
Also known as
- Morbidity and mortality conference
- M&M
- Blameless post-incident review
- Learning review
Categories
- Accessibility
- Human Factors & Error