Most turnaround stories skip the hardest part. They tell you what the company pivoted from and what it pivoted to, and they leave out the months in between when the old business was still paying the bills, the new one wasn’t paying anything, and the founder was running both with the same hands and the same bank account. That middle stretch is where most pivots actually die, not because the new idea was bad, but because the operator ran out of money, energy, or nerve before it had a chance to prove itself.

The solution isn’t to jump. Quitting the old business on the strength of a promising idea is how people end up with no income and a half-built experiment. But the opposite is just as dangerous. Keeping the old business running “for now,” with no end date, means the new thing gets whatever attention is left over, which is usually none, and a year later you’re still doing the work you wanted to leave, with a side project that never got a fair test.

The fix is a deadline, written down, for the old business. Not a deadline to shut it down, necessarily, but a date by which you’ll decide. Pick a horizon that’s long enough for the new thing to show a real signal and short enough to feel like pressure. For most solo operators that’s somewhere between three and six months. Then write, in one sentence, what the new business has to show by that date to earn the next stretch of your attention: a certain number of paying customers, a certain amount of revenue, a certain number of people who said yes without being your friends. Be specific. “Some traction” is not a threshold, it’s an excuse to keep going or stop based on how you feel that week.

Between now and the date, the old business has one job: fund the experiment. Stop treating it as the thing you’re building and start treating it as the thing that’s paying for the thing you’re building. That means taking the work that pays and turning down the work that merely keeps you busy. It means no new long-term commitments in the old business, no fresh investment in it, no redesign of its website. It’s a bridge, and you don’t renovate a bridge you’re planning to cross.

Then protect a fixed block for the new thing, and make it the first block, not the last. If the experiment gets whatever hours survive the day, it will get nothing, because the old business is better at generating urgency. Two mornings a week is enough to run a real test if the mornings are actually protected. Two “when I get a chance” evenings a week is enough to run nothing.

When the date arrives, look at the sentence you wrote and answer it honestly. If the new thing cleared the bar, it earns more of you, and you can start planning the real transition. If it didn’t, that’s a result too, and a far cheaper one than the version where you found out after two years. Either way you’ve replaced the slow bleed with a decision, and a decision, even a disappointing one, is something a business can survive.