Decision Fatigue is Killing Your Team's Productivity
Most teams feel the drain of endless choices: priorities, approvals, channel pings, and tiny process calls that add up. The popular claim that we each make “35,000 decisions per day” is widely repeated, but we could not find a primary, peer‑reviewed source for that exact figure. Still, the broader story is well supported: micro‑decisions, interruptions, and context switching erode execution capacity.
This post summarizes what the research does and doesn’t say, then offers practical, system‑level changes product and engineering orgs can implement—without hype.
#What the research actually says (and what it doesn’t)
• Decision fatigue as a concept was popularized in behavioral science and mainstream writing (e.g., Baumeister & Tierney, 2011). The idea: repeated acts of choosing and self‑control can reduce subsequent decision quality.• Evidence is mixed. A large critique and meta‑analysis found signs of publication bias and smaller‑than‑claimed effects for ego depletion (Carter & McCullough, 2014; Carter et al., 2015). Translation: be cautious about sweeping claims.• Real‑world decision patterns can drift over time. A well‑known study of Israeli parole boards found favorable rulings were more common earlier in the day and after breaks (Danziger et al., 2011). The mechanism is debated, but the pattern warns that decision environments matter.• Interruptions and context switching are costly. Knowledge workers lose significant time recovering after disruptions; one field study reported around 23 minutes to resume the original task after an interruption (Mark et al., 2008).• Communication load crowds out focused work. Microsoft’s 2023 Work Trend Index reported people spend the majority of their time communicating rather than creating, indicating substantial cognitive overhead in modern collaboration systems (Microsoft, 2023).
Bottom line: whether or not a single grand statistic captures it, the practical costs of fragmented attention and constant micro‑choices are real enough to address with better systems.
#Where decision fatigue shows up in product teams
• Slack/Teams spiral: ambiguous asks in public channels lead to long back‑and‑forth threads, re‑asking for context, and parallel DMs that create inconsistent decisions.• Unbounded approvals: one‑off exceptions and ad‑hoc approver lists force repeated triage for similar requests.• Meetings by default: recurring status meetings substitute for clear async decision mechanisms, consuming time and diluting accountability.• Bikeshedding: small, low‑impact choices consume disproportionate debate because there’s no agreed default or threshold for “good enough.”
#Practical systems to reduce decision load
Treat decision fatigue as a design problem for your operating system.
1) Make good defaults the norm
• Publish defaults for common calls (priority labels, rollout steps, incident severities, acceptable risk ranges). Defaults reduce choice complexity and decision variance (Thaler & Sunstein, 2008).• Pair defaults with explicit “when to override” criteria.
2) Use lightweight checklists for repeatable work
• For releases, incidents, handoffs, and onboarding, checklists cut omissions and debate about basics (Gawande, 2010).• Keep them short. If a checklist grows unwieldy, split it.
3) Separate reversible vs. irreversible decisions
• Document when to ship fast (reversible, low‑blast‑radius) vs. when to seek a deeper review (one‑way door). A two‑tier path prevents over‑deliberating routine choices.
4) Move debates into structured async proposals
• Use concise RFC or ADR templates with problem, options, trade‑offs, and decision. Time‑box feedback. Default to accept if no blocking concern emerges by the deadline.• Keep decisions searchable in a log so you don’t re‑decide the same topics.
5) Create clear intake and routing
• For common requests (access, data pulls, minor exceptions), use forms or short templates to capture must‑have info and route to the right owner automatically.• Publish SLAs. Knowing “what happens next” reduces escalations and DMs.
6) Calibrate meeting protocols
• Require a written brief for any decision meeting: context, options, proposal, owners, and a time‑boxed agenda.• Reserve synchronous time for contention, not for broadcasting.
#Where AI mediation helps (without the hype)
Used carefully, AI can reduce decision friction by:
• Summarizing long threads into a crisp brief with options and trade‑offs.• Nudging contributors to fill missing context using your team’s templates.• Suggesting relevant prior decisions from your log to avoid re‑litigation.• Auto‑routing routine requests to the correct queue, escalating only exceptions.
Keep humans in the loop for policy, risk, and ambiguous trade‑offs. Use AI to standardize inputs and surface context, not to make irreversible calls.
#Caveats
• Effects vary by team and workload. Treat the above as hypotheses to test.• The decision‑fatigue literature has mixed findings; design changes should be grounded in your data and constraints.
#References
• Baumeister, R. F., & Tierney, J. (2011). Willpower: Rediscovering the Greatest Human Strength. https://www.penguinrandomhouse.com/books/307839/willpower-by-roy-f-baumeister-and-john-tierney/• Carter, E. C., & McCullough, M. E. (2014). Publication bias and the limited evidence for ego depletion effects. Frontiers in Psychology. https://www.frontiersin.org/articles/10.3389/fpsyg.2014.00823/full• Carter, E. C., Kofler, L. M., Forster, D. E., & McCullough, M. E. (2015). A series of meta-analytic tests of the depletion effect. Journal of Experimental Psychology: General. https://doi.org/10.1037/xge0000086• Danziger, S., Levav, J., & Avnaim-Pesso, L. (2011). Extraneous factors in judicial decisions. PNAS. https://www.pnas.org/doi/10.1073/pnas.1018033108• Mark, G., Gonzalez, V. M., & Harris, J. (2008). The cost of interrupted work: More speed and stress. CHI. (Summary overview) https://www.ics.uci.edu/~gmark/chi08-mark.pdf• Microsoft Work Trend Index 2023: Will AI Fix Work? https://www.microsoft.com/en-us/worklab/work-trend-index/will-ai-fix-work• Thaler, R. H., & Sunstein, C. R. (2008). Nudge. https://www.penguinrandomhouse.com/books/171682/nudge-by-richard-h-thaler-and-cass-r-sunstein/• Architecture Decision Records (ADR) template. https://github.com/joelparkerhenderson/architecture_decision_record• Gawande, A. (2010). The Checklist Manifesto. https://www.penguinrandomhouse.com/books/56367/the-checklist-manifesto-by-atul-gawande/