What dashboards don't tell you
summary
Argues that self‑perception matters as much as telemetry and that dashboards often surface vanity metrics that can be gamed (Goodhart’s law and simple python scripts closing PRs). Shows overlaying incidents onto pull‑request activity to reveal hidden tension, and recommends starting with a survey before picking metrics. Points to research like SPACE productivity metrics, mentions personal use of rescue time, and suggests qualitative tooling such as dx that avoids importing github or jira data.
AI-written summary · about
chapters (4)
- 0:00Start
- 1:37Performance and productivity are multi-dimensional
- 8:10Avoid vanity metrics with tension
- 9:11Start with self-perception
related talks
37mA systems view of developer productivity
25mBeyond Dashboards: Cultivating the Observability Mindset
26mMetrics and Modeling – Helping Teams See How to Improve
45mCreating & Maintaining Impactful Dashboards'
35mWho Needs Dashboards?
30mDevEx and Productivity Metrics on Backstage - Do's and Don'ts
21mDashboards and Runbooks: Scrapbooking for Engineers
44mIterative metrics, dashboards and monitors