Office of Small Failures
Tiny products. Public tests. Useful mistakes. Nothing gets promoted just because somebody built it.
Office of Small Failures is a tiny internet product studio. Small things get built, exposed to actual use, reviewed, and either kept, repaired or killed. The files below are the experiments currently allowed to exist.
Open case files
Four experiments / one current priorityWorthDoing
A small SEO decision machine. Put in a task, likely upside, effort and uncertainty. Get back a commercial case and a verdict on whether it deserves to survive the backlog.
Should I Automate This?
A small decision tool for recurring work. Work out whether something deserves automation, AI assistance, ordinary software or a human being.
AI Mute
An experiment in reducing AI-content fatigue by hiding the stuff you no longer want to hear about.
Signal Sample
A deliberately sampled way to measure visibility across AI-search surfaces without pretending there is a universal AI ranking. Sampled evidence, not a ranking.
Office policy
Failure is cheaper when the thing is still tiny.
Promotion criteria
Useful things get another week. Everything else gets paperwork.
How a case file survives
Very small methodology departmentBuild enough to test the idea. Put it in front of real use. Review the evidence. Give useful things more time; repair specific problems; close the rest before they become infrastructure.
Submit a problem
Got a specific digital or work annoyance that deserves a tiny experiment? Send it to the Office intake desk. Specific irritating problems are much more useful than startup ideas.