One Engineer's Year of Screen Logging Finds Weather Outpaced Sleep as Productivity Signal

According to a blog post from donethat, one engineer spent a full year running the company’s screen-monitoring tool on their own machine and logged correlations between work activity and external factors. The finding that struck them: weather patterns appeared to predict their productive output more reliably than sleep duration did—a counterintuitive result from a personal dogfooding exercise.

A Year of Self-Directed Monitoring

The engineer used donethat’s own product to track their screen activity continuously over 12 months. Rather than treating the tool as purely a work-logging instrument, they cross-referenced their screen-time patterns with lifestyle and environmental variables, including sleep hours and daily weather. The experiment was not conducted under controlled research conditions; instead, it represents authentic product usage and the kind of naturalistic observation that often surfaces insights missed in lab studies.

Weather Emerged as the Stronger Signal

Among the variables tracked, weather conditions showed a more consistent relationship with the engineer’s productive screen time than sleep metrics did. While sleep duration is conventionally treated as a primary lever for cognitive performance, this personal observation suggests that environmental factors—cloud cover, temperature, precipitation, or seasonal light patterns—may exert measurable influence on day-to-day work output at least for this individual.

Why This Matters

For people experimenting with personal productivity tracking, this anecdotal finding suggests that adding weather or environmental data to your existing sleep-and-work datasets could be worth testing. The relationship between environment and focus is not new to chronobiology or organizational psychology, but individual-level logging tools like donethat now make it feasible for workers to validate whether the correlation holds in their own routines. The real value is not in generalizing from one person’s year, but in lowering the friction to run your own single-subject experiment.