You don't have to keep doing things that don't work

I’m working on a new project for founders and want to be sure I’m building something you actually need. If you want more growth, meetings, revenue, or customers–click the button!

The first 10 customers are the hardest (click the button 👆️). There are different ways to get to those 10 customers, some are better than others, but fundamentally what you are looking to figure out is whether the market wants your product, and whether they’ll pay for it (or, at least, whether you can make money). It’s an experiment, or more accurately, several conjoined experiments. The first of many.

Many founders and their teams get so attached to their experiments, they forget that they are, in fact, just experiments. The whole enterprise is, actually, an experiment. The founders that succeed realize they don’t have to keep doing things that don’t work–not only the smaller experiments, but the whole company.

It’s a fine line between staying committed to your vision and being in denial about what isn’t working. The more you invest the more you are invested. Pivoting away from a product with 5 users is harder–but no different–than shutting down a product with 20,000 users. Continuing to invest time and resources into things that don’t work—even marginally—distracts you from doubling down on what does.

The Expensify example

Expensify is a public company today. I met the founder, David Barrett, when I was at TripIt. He had just launched an expense management product–in the form of a credit card. You put your expenses on the card, and it had nifty software that turned it into expense reports.

Nobody liked the credit card. If I remember right it was TechCrunch Disrupt, one of the first ones, and the reaction was somewhere between tepid and aggressively negative, especially from investors. Adoption was slow. People were happy with their Amex, or other card, and had zero interest in the Expensify card.

But they loved the software.

David quickly pivoted, killing the card and doubling down on the software. The company went public in 2021, around 12 years after killing the card experiment.

Being wrong sucks

Sticking with something that isn’t working is an easy trap to fall into. Letting go feels like admitting failure after pouring months, or years, into something and telling everyone that you have things figured out. The bigger the investment the bigger the attachment.

So many founders know they need to shut down something–whether that’s letting go an underperforming exec, a product, a major marketing initiative, or even their company–but starting over feels like a step backwards, and they’d rather crawl forwards.

Evidence that supports your perspective is everywhere, and your brain gravitates towards it and emphasizes it. That one new customer, one successful pilot–in a sea of churn and failed pilots–might keep the failed experiment alive. This is called confirmation bias, and it’s how we’re wired as humans. Even if you’re aware of it, you’ll still succumb. Damned subconscious!

Perhaps the worst part of being wrong can be the pressure from other people. It’s hard to tell your team, investors, or customers that you were wrong, and change the plan, without also feeling like you’re letting them down. This is particularly acute in YC, where founders frequently get sucked into an optics focused culture.

Then the reality of actually shutting something down–whether a product, or a marketing program, or an entire company–truly sucks. It is perhaps the least pleasant of all work: who wants to answer angry customer emails? Tell investors they lost their money? Tell employees that the company is going in a different direction, and they have to go find a job somewhere else? We are hardwired to hate this kind of work, so like teenagers we do all we can to avoid it.

A cautionary tale

We usually hear about the successes. Slack, pivoting from a gaming project. Twitter emerging out of the ashes of Odeo, a failed podcasting platform (on that subject, had a great conversation about that pivot with Tony Stubblebine on the podcast). What the public record is missing are examples of when you cling to your experiment, and don’t pivot.

I knew a founder who had launched a product with a ton of PR. His knack for press was uncanny, and it was a hot mobile app in a hot space back in the early 2010s. But retention was literally 0–nobody wanted to use the app a second time. But he was somehow able to get it featured by Apple anyway. Big tile in the app store in the social networking category, right next to NextDoor. Company-making PR, 100% founder hustle.

Through friends at NextDoor I knew their app was getting 20,000+ downloads every day because of that tile. This app was getting around 1,000. And exactly 1,000 of those people churned out after 1 session. People clearly didn’t want what he was selling. What did he do?

He hired a team to build an Android version.

Obviously that wouldn’t solve anything. If Apple users didn’t like it, Android users weren’t going to like it either. About six months later I helped him get acquihired by a large tech company. It didn’t have to be that way. He had plenty of money and supportive investors, good product design skills and a knack for PR. There was, maybe, a path to success. But instead, he clung to a failed experiment. It felt better not to be wrong than admit it didn’t work and try something different.

You can be right at the wrong time

If we go back to Expensify, David was right–there was, eventually, a market for a credit card tied to expense software. Expensify launched one a few years ago. Brex also launched one targeting founders and it’s one of the more successful startups of our time.

But that doesn’t change the fact that the credit card experiment failed. There are many, many reasons one might keep at a product that isn’t working. But the best founders are honest with themselves, figure out how to have grit and stubbornness only to a point–while letting go of what isn’t serving them. Then they come back to it when the time is right.

To shut down or to experiment

I talked to a founder yesterday about getting acquihired. She was sent by a prominent accelerator because she was stuck. Her cofounder quit to take a job. Their pilots hadn’t converted. I’ve been there. It sucks.

But the company still had cash in the bank. The pilots had been design partners and helped her build out a pretty useful sounding product. There were several areas where execution had maybe missed–for example, selling time savings rather than revenue growth or loss aversion. Of course, if she wants to stop there’s no shame in that. But I thought, fine, those experiments didn’t work. What did she learn? Could she admit she was wrong in some places, dig through the data, and run another experiment or two? Maybe success is in there.

During the dark times, it’s easy to forget how great it is when your experiments work. Warmly’s founder (and S2S reader) Max Greenwald wrote a great piece this week in Kyle Polar’s Growth Unhinged newsletter about the experiments they ran to get from $1 million to $3 million in ARR. It’s worth a read. Many of their experiments worked, and they tripled their revenue. Kudos. That’s what you’re going for.

When you’re stuck on something that isn’t working, ask yourself or your team: “Am I holding on to this because I believe it really has potential, or because I’m afraid to let it go?” The sooner you redirect your focus, the sooner you’ll uncover what does work. If the potential is there, keep going. But if it isn’t, it’s ok to admit you’re wrong and pivot. All the great companies did it at least once. The best, like Warmly, do it all the time.

If you read this far you’ve got the 30 seconds it will take to answer 7 questions about my new project–that could be the experiment that gets you to $1mm ARR or whatever your next milestone is. Click the button!