Most startups don't fail because of bad timing, or tough competition. They fail because the founders committed one of four sins — and usually more than one. These are the undeniable, hard-won lessons that most founders still ignore because they think the rules are different in their case.
They are not.
Sin 1: Not Solving a Worthy Problem
A worthy problem is one that people actually have. You know it's a worthy problem because when you ask them if they'd pay to have it solved, the answer is "hell yes!" Usually they've tried other things that didn't work. The demand for your product existed before you arrived.
When we started Instacart, the problem was obvious. People shop for groceries weekly and it's a huge chore. We thought if we could make it cheap enough, those people would want our grocery delivery service. The demand had been there for decades.
What's it look like to solve a not worthy problem? Typically founders fall in love with a technology or the idea of being a founder, and start with "wouldn't it be cool if?"
Markets are efficient, so it's hard to find an unsolved worthy problem. But it's necessary. If you haven't found a worthy problem, you shouldn't do much else until you do.
Sin 2: Not Talking to Customers, or Not Asking the Right Questions
Customer development is not just how you get the insight into what to build in the first place, but a continuous process you run while building your company. Every time you change the product, you must check back with customers.
When interviewing customers, you're not looking to average their feedback — you're mining for insights. Like nuggets of gold, these insights are unevenly distributed across conversations.
If you do ten customer conversations, the first eight may yield nothing new. The ninth customer might say something that catches your attention. Then you ask the tenth person about that thing — and they tell you something that changes your perspective on the market. Now you have a customer insight that you can work with to shape your product. Don't give up after talking to the first nine customers, since the insight is often found after many similar and unproductive conversations with the wrong customers.
To uncover insights you must ask the right questions. Start open ended before leading the customer. "If you could wave a magic wand, what would you ask us to build?", "What is your spend in this category?" "Is solving this one of your top 3 priorities?" Then, and only then, show them what you're working on and figure out if it would be valuable to them.
Sin 3: Not Launching, or Not Iterating Fast Enough
People wait too long to launch. They're scared. They're perfectionists. They don't want to burn a customer with a half-baked product. Deep down, they're afraid their product might fail.
It will fail. The art of finding product-market fit is figuring out how your product will fail, deciding how to fix it, and then repeating the cycle fast enough to reach product-market fit before your company dies.
You cannot run this process inside the safe container of your company and friends. You must launch to real customers. The point of launching is feedback — and feedback from real customers is one thing an AI model cannot give you.
After you launch, things might seem like they're going well. You might even taste early product market fit. But the job is not done. Now you need to collect these early signals, lean in and double down.
Strong product market fit feels like a wind being blown into your sail. Selling your product gets easier. Customers refer other customers. You can't keep up with demand. When this starts to happen, it's the right time to iterate even faster, reshaping your product around what's working, and jettisoning the pieces that are slowing you down.
Sin 4: Not Being All In
The fourth sin is not being all in. You won't be successful unless you're irrationally excited about what you're doing, and you won't nail PMF unless you focus all of your energy on one product.
Building a startup is a 7 day a week job — it's not for everyone. You cannot build a startup from vacation — even with 100 AI agents running 24/7. As glamorous as it might sound, you likely can't do it without a team either.
Your team needs to be all in too. You also cannot expect your team to outwork you. No one will exceed your bar for quality or your bar for urgency so you have to model both yourself, every day, in how you behave. The team will calibrate to whatever you do.
For the second time founders out there: the gods of product-market fit don't care what you did before. Even more than a first time founder, you need to focus all your energy on your startup and approach it with a beginner's mindset.
You must…
Start with a worthy problem — talk to your customers — ask the right questions — launch and iterate — and be all in. You'll give yourself the best shot at finding product market fit, and be on your way to building an epic company!