The decision that keeps me up I've watched people running their own thing make this call wrong more often than right. They either kill something that was three months away from working, or they pour another year into something that died two years ago. Neither one is forgivable if you had a way to know. The thing is, you do have a way to know. It's not magic. It's not intuition, though people always want to call it that. It's just a framework applied to actual data instead of hope. I see this happen across different types of work. Someone's built a service that isn't gaining traction. Someone's got a product that's technically sound but nobody's buying it. Someone's team is burning out maintaining something that doesn't move the needle. The question is always the same: do we keep going or do we change direction? Right now, with markets showing volatility over everything from geopolitical risk to housing costs, I'm getting asked this more. People are watching margins tighten, customer behaviour shift, and they're wondering if it's time to adapt or if they're just being spooked by noise. The persevere trap Most people get this wrong by staying too long. I say this because I've seen it more than the other way around. Perseverance is a story we tell about success. Every successful person talks about how they nearly quit but didn't. So when you're struggling, you think that's just the story you're supposed to be in. You're supposed to push through. You're supposed to believe. But belief without evidence is just expensive hope. I've worked with people who've sunk 18 months into something where the core assumption was broken from month two. They knew it. They just didn't have a way to say it out loud, so they kept moving. The persevere trap gets worse when you've already invested. You've spent money, time, your reputation. Walking away feels like admitting defeat. So you invest more to justify the investment. Behavioural economists call this sunk cost fallacy. I call it watching people spend their last 12 months trying to save the previous 12 months. You see this in the market right now. Mortgage rates are falling, which feels like relief, but underneath that's the reality that some people bet wrong on where rates would go and now they're adjusting. Young people are living with their parents longer because the maths got worse, not because they wanted to stay home. Businesses that built their model on certain assumptions are discovering those assumptions moved. The pivot trap The other way to get this wrong is faster but feels less painful, which is why people do it. You're not getting traction. So you pivot. You change the target customer or the product or the messaging. Three months later, that doesn't work either. So you pivot again. I've seen people do this 5 times in 18 months. They feel like they're being responsive and agile. What they're actually doing is never testing anything long enough to get real signal. You need time to know if something works. Not years. But not weeks either. The pivot trap is seductive because it feels like you're doing something. You're not failing, you're iterating. You're not stuck, you're adapting. Except if you pivot every time you hit friction, you never get through the friction that's actually fixable. The framework that works Here's what I actually do when someone asks me this. First, separate your assumption from your execution. Write down what you believed would be true. Then write down what you've actually learned. These are rarely the same thing. Second, identify which part is broken. Is it the market that doesn't want this? Is it that the market wants it but you're not reaching them? Is it that you can't build it profitably? Is it that you can't build it at all? These are completely different problems with completely different solutions. Third, ask whether you can fix the broken part with reasonable effort. If your assumption was wrong, you pivot. If your execution was wrong, you persevere and fix it. If it's unfixable within realistic constraints, you stop. I use a Decision Matrix for this with clients at ALIRA. You're rating things like: how confident are we in the core assumption now? What's the evidence? How much effort to test a fix versus how much effort to pivot? What does our runway actually allow? Then you're not deciding on emotion, you're deciding on what the data shows. Fourth, set a clear stop date before you start. Not a vague "we'll see how it goes." A specific date. 8 weeks. 12 weeks. A specific metric you need to hit. If you hit it, you keep going. If you don't, you decide. This removes the temptation to keep extending because you feel like you're close. What actually matters Persevere when your core assumption is still valid but your execution needs fixing. When customers want what you're building but you're not reaching them fast enough or cheaply enough or clearly enough. When the problem is you, not the market. Pivot when your core assumption is broken. When you've tested it properly and the market is telling you something different from what you expected. When you can clearly identify that the thing you set out to build isn't what people actually need. Stop when it's unfixable. When you can't reach the market, can't build it profitably, or can't do it with the resources you have. This is the one people avoid saying out loud. But sometimes the right decision is to stop and do something else. The difference between these three is usually visible in the data if you look at it honestly. Customer feedback. Usage metrics. Conversion rates. How hard you have to work to get each sale. Whether people come back. Whether your unit economics work. These tell you which category you're in. What to do this week Write down the three core assumptions you made when you started what you're running now. Not the nice version. The real assumptions about who would buy, why they'd buy it, and how you'd reach them. Next to each one, write what you've actually learned. Be specific. If your assumption was "small agencies will pay £500 per month for this," write down what you've actually learned about pricing. If it was "we'll reach people through LinkedIn," write what you've learned about that channel. Then answer one question honestly: is the gap between assumption and reality something you can fix with effort, or is it something that means you're building the wrong thing? That answer tells you whether to push or change direction.