Kwame's broken signup at 11:47 PM. The Lagos customer decides what she tells two colleagues by morning.

What this post answers up front

Yes, six months of runway is solvable, and yes, a broken signup flow will quietly kill you. But they are not the same kind of problem, and mistaking one for the other is how founders burn both. One is a date on the calendar with consequences weeks away. The other is happening to a customer at 11:47 PM tonight, and it decides what she tells two colleagues tomorrow morning.

The application and the signup

Kwame had both deadlines open on his laptop at 11:47 PM, two tabs, and he kept clicking between them like the answer was in the switching.

The first tab was the overseas contract. Six months of runway left, and here was a company in Berlin offering enough money to extend that to eighteen. The work was adjacent to what Asenda does, not the same thing, but adjacent had never paid this well before. The email sat there, read twice, unanswered, because answering it felt like making a decision about the next two years of his life.

The second tab was the onboarding flow. A pilot customer in Lagos had sent a screen recording of what happened when she tried to sign up for the product. Four minutes of footage, mostly her clicking the same button and watching nothing happen. She had not written a single word of complaint. She had just recorded the moment and sent it, which somehow made it worse. The signing up did not work, and she had bothered to document it.

He had told himself all week that the contract was the strategic decision and the signup was a bug. One was a chess move. The other was mopping the floor. Every founder he admired was doing strategy. Nobody was getting featured for fixing a button.

The trap in the distinction

Here is what took him another two weeks to see. The contract was a decision about the future, and it could wait until tomorrow, or next week, or the week after, without the future changing by a single day. The offer was not going to evaporate because he slept on it. The signup, though, was happening in real time. Every hour it stayed broken, the woman in Lagos was telling herself the product did not work, and by extension, that he did not work, and by extension that the whole thing was a demo wearing a trench coat. She was not going to email him about it. She was going to tell two other founders at a meetup, the way you warn people about a pothole.

The contract was a decision he could make well. The signup was a cost he was already paying.

What actually changes what customers experience

The trick that broke the logjam was not discipline or a productivity system. It was asking which of the two tabs, if resolved tonight, would change what a customer felt tomorrow morning.

The contract, resolved tonight, would change nothing. Not one user would notice. The signup, resolved tonight, meant the woman in Lagos could get in, see the value, and have a reason to come back. One of these had a direct line to the thing he was actually building. The other was a paycheck with a long delay.

He fixed the signup first. It took him four hours, most of it spent discovering that the failure was not the button at all but a validation rule that rejected every email address from two of the largest Nigerian providers. His own product, built by him, in a market he claimed to serve, had been silently turning away the exact people he was trying to reach. The woman in Lagos was not a fluke. She was the rule.

The contract he answered the next morning, and it was a better answer for having waited. Not because he was fresher. Because fixing the signup had reminded him which market he was actually building for, and the answer to the Berlin offer changed accordingly. He took the contract, but on terms that protected the Lagos roadmap, because he now knew what the roadmap was.

The decision rule that survived

Kwame keeps a version of this in his head now, and it has outlasted the specific panic of that night.

When two things are both urgent, sort them by what a customer will feel tomorrow morning. Not by what looks more strategic. Not by what will look good in a monthly report. The thing with a direct line to a user's experience today outranks the thing with a line to your bank account in six months, almost every time. The one that can wait can always wait. The one that is happening to a real person right now cannot.

This is the same call as the one in [Jonas's offer]( /blog/jonas-s-offer-paid-the-bills-saying-yes-would-have-killed-the-market-research-he-still-needed-f900a567/ ), where the money and the research pointed in opposite directions. The pattern keeps showing up because it is the actual texture of building on limited runway: the choices that pay you and the choices that build the product are often different choices, and only one of them has a customer attached to it today.

He still cannot tell you which tab he opened first at 11:47 PM. But he can tell you which one he closes, now, when both are on screen. The woman in Lagos never wrote him a thank-you note. She did not need to. She just signed up, and used the thing, and came back the next week.