The founder's quest
capability: Method · last revised August 2026
The jobs page is the map. This is the territory: the full written version, from my own runs. The quest runs from the day you commit to an idea until demand becomes repeatable without heroics — product-market fit, which is pull: retention without persuasion, demand you didn't manufacture. The quest is measured in evidence, not months. Its length is unknowable from inside, and that matters more than it sounds: you might be at mile 2 or mile 20, and nothing on your dashboard says which.
Learn and validate
The defining job, and the one I didn't write down my first time through. Name, out loud, the riskiest assumption you are standing on. Run the cheapest experiment that could kill it. Killing is not loss; it is polishing. A rough idea is mostly noise — untested assumptions, borrowed features, imagined customers — and each honest experiment removes what the evidence will not support. What survives is the essence: the thing people actually want. Less is more is not an aesthetic here; it is the mechanism, and it is why speed matters — every week a false assumption survives, it costs attention the essence needed. Talk to real users every week — to listen, not to pitch. Once there is first signal, selling becomes the sharpest experiment available: nothing exposes a value proposition like asking someone to pay. You sell for truth long before you sell for revenue. And hold one test above the rest: if a week ends and nothing you learned changes what you do Monday, you weren't learning. You were moving.
Decide and execute
The engine: decide the single most valuable next action — singular, not a list. Execute it before the conviction behind it decays. Measure whether it moved a real needle, meaning a user behaviour, not your own activity. Feed the result into the next decision. The loop is identical at day five and day five hundred; what a long quest changes is the cost of each turn. Which reframes runway: runway is not money. It is the number of learning cycles you can still afford. A founder with twenty cycles and slow learning loses to a founder with eight cycles and fast ones — so the job is to maximise cycles remaining, from either end: extend the cash, or make each experiment cheaper.
Sustain conviction and relationships
The job nobody writes down, and on a long quest, the one that decides survival. Carry the belief: hold conviction with zero external validation, and re-hold it after every false positive — an enthusiastic pilot that goes nowhere costs more than silence, because it recalibrates your sense of remaining distance, wrongly. Protect the self: separate identity from idea so the idea can die without taking you with it; pace for unknown distance — run like it could end in weeks, structure like it could take years; and learn to tell fear-driven avoidance from legitimate rest, because I got that wrong in both directions. Protect the people: keep the few who matter genuinely committed for the full unknowable length of the road, and do not let the people you love quietly pay the price. No customer will ever complain when you miss dinner. This job has no external forcing function, which is exactly why it must be written down.
The two loops
The engine — decide, execute, measure, decide again — has a shadow: decision paralysis, then work-avoidance dressed as progress, then emotional decisions made under pressure, then deeper paralysis. I did not fall into the shadow loop from laziness. I fell in every time unvalidated assumptions piled up faster than I resolved them: when too many open questions stack, every decision feels load-bearing, so none feels safe. A long quest adds a second trigger — milestone starvation. Humans need finish lines and product-market fit provides no interim ones. The way back was never willpower. It was shrinking the next decision until it was small enough to survive being wrong, and manufacturing honest checkpoints: named, dated, falsifiable. Artificial finish lines are not self-deception if the criteria are honest. They are pacing. Paralysis is a symptom of assumption debt, not a character flaw.
The pains, ranked
Ranked by severity, frequency, and how badly existing advice failed me. First: not knowing whether I was learning or just moving — my activity was excellent and my learning was absent, and no productivity system caught it because every system measured activity. Second: not knowing where I was on the road — every metric measured speed, none measured remaining distance, and that gap parents both burnout and conviction decay. Third: the shadow loop, compounding, because each emotional decision created new assumption debt. Fourth: being alone with the belief — no one sees it, there is no metric for it, and low morale is only its symptom. Beneath all four sits one root condition: clarity of why. With it, everything above is survivable. Without it, everything metastasises.
How to use this
Read the pains and check your body, not your head — the ones that are yours will wince. Run the Friday test for three weeks. And watch for your own flip triggers: mine were assumption debt and milestone starvation; yours may differ, but the shadow loop almost certainly has a mechanical trigger, and finding it beats every motivational fix.