Why 720 hours is the right delivery window
Six months is a hedge, not an estimate. Two weeks is bravado. Thirty days is the only window that respects both founder runway and engineering reality.
There’s a number-format reason kaedax says “720 hours” instead of “30 days.” Time pressure hits differently when it’s denominated in something you can spend.
But the deeper reason is calibration. Most agencies quote in months because months are an emotional unit — they buy distance from the conversation. Six months is the polite way to say “we don’t actually know.” Two weeks is the cowboy version of the same.
Thirty days — 720 hours — is short enough that you can’t pretend, and long enough that you have to plan. It’s the smallest window where an agent loop can productively repeat. It’s the largest window where a bootstrapped founder doesn’t start hemorrhaging momentum.
What 720 hours buys
- One real product cycle. Scope, build, polish, ship — sequentially, not concurrently. You can’t run “discovery” in parallel with “delivery” and have either be good.
- Six iterations of the agent loop. Each agent gets multiple passes at its part of the spec. By cycle three, it knows your codebase better than a senior engineer would on day one.
- Time for the unsexy week. Days 22-27 are where the actual product gets made. Most shops skip this and pretend the demo is the product.
What 720 hours costs
Honesty. We can’t take builds we can’t fit. We turn away maybe 30% of inquiries on the scope call — usually founders who want “a small platform” or “a quick rebuild of our backend.” Both of those are 1,800-hour jobs at best, and pretending otherwise would be the trick that makes 720h sound impressive but feel impossible.
The window is the discipline. We’re protective of it.