When to Kill Your "Working" Product and Start Over

There's a painful truth that most tech founders don't want to hear: sometimes a product that's "working" still needs to die.

I recently interviewed Pablo Srugo on my Product Driven podcast to discuss this exact challenge.

Pablo serves as Partner at Mistral Venture Partners and previously co-founded Gymtrack as Co-CEO. With experience on both sides of the table—as founder and now VC—and hosting the Product Market Fit podcast, Pablo has seen this pattern play out hundreds of times.

One story he shared stuck with me. A founder with $70K ARR going through Y Combinator decided to completely shut down his product, fire all his customers, and start over. Why? Because despite having revenue, he didn't feel the market pull that signals true product-market fit.

One year later, this founder had built an entirely new AI product that went from $0 to $2M ARR.

It's a perfect illustration of a decision every technical founder eventually faces: when do you kill a product that's "kind of working" but clearly not achieving escape velocity?

The Gray Zone That Kills Companies

If your product really isn't working at all, you're probably going to keep iterating.

If your product's totally taking off, great.

The challenge is that gray area where you have some pull, some traction, but you don't have true product market fit.

I lived through this at Stackify. We had respectable metrics—about 30% demo-to-close rate, solid ARR growth—but we never achieved that magical market pull where growth becomes inevitable. We were always pushing.

This gray zone is dangerous for three reasons:

It's financially sustainable enough to survive, but not thrive.

You're making just enough money to keep the lights on. Every month, you're covering your AWS bills, paying your small team (though maybe not what they deserve), and maybe even drawing a modest salary yourself. But there's never enough left for meaningful growth investments.

This is startup purgatory.

The cruel part is that you're not failing obviously enough to force a decision. You're bringing in new customers... just not enough of them. Your MRR chart shows growth... just not the hockey stick you promised investors or yourself. You find yourself constantly adjusting forecasts downward, explaining why "next quarter" will finally show the breakthrough.

I've been there. At Stackify, we were growing, but never growing fast enough. We'd celebrate hitting $100K MRR, then $200K, but each milestone took longer than projected. It's like running a marathon where someone keeps moving the finish line.

The emotional toll is brutal.

You keep the positive founder face on during team meetings, but privately wonder if this is as good as it gets. You start redefining success in your own mind, convincing yourself that "steady growth" was the goal all along. Meanwhile, you're watching other companies in adjacent spaces raise massive rounds and scale exponentially.

The sunk cost fallacy becomes overwhelming.

Years of your life. Millions in investment. Thousands of git commits. Team members who joined because they believed in your vision. Early customers who took a chance on you. All of these create immense psychological weight against pivoting.

"How can I tell my team we're changing direction after everything they've built?"

"What will our investors say if we admit this approach isn't working?"

"We've spent 18 months perfecting this architecture—we can't just throw it away."

These thoughts consume you. Every all-hands meeting where you've confidently presented "the roadmap" makes it harder to admit uncertainty. Every investor update where you've defended your strategy makes it almost impossible to question it.

The weight of past decisions literally changes how you perceive new information. You start unconsciously filtering customer feedback, amplifying positive signals and dismissing warning signs. Data that contradicts your core assumptions gets labeled as "anomalies" or "not our ideal customer segment."

I've caught myself doing this at multiple companies. At GymTrack, Pablo described working on their analytics platform for a year plus despite signs it wasn't solving a top pain point—because they had a false sense of validation from initial customer enthusiasm. That validation bias is insidious.

The opportunity cost is invisible but massive.

This might be the most dangerous aspect of all—because you literally cannot see what you're missing.

Every day spent optimizing a product with mediocre market fit is a day not spent building something people are desperate for. Every developer hour invested in refining features nobody truly needs is an hour not invested in discovering what customers would actually pull from your hands.

When you're heads-down executing on a "working" product, you're not exploring alternative ideas that might have 10x the potential. You're not having the customer conversations that could reveal breakthrough insights. You're not experimenting with radically different approaches.

The catch-22 is that you only realize this opportunity cost in retrospect. Pablo told the story of a founder who shut down a product doing $70K ARR, completely started over, and built a new AI product that went from $0 to $2M ARR in a year. The opportunity cost of sticking with the first product would have been enormous—but completely invisible until he made the leap.

I've felt this acutely. After successfully exiting VinSolutions, I realized how much time we'd spent on products and features that never mattered to our ultimate success. If we'd pivoted sooner from our original car photo product to our CRM focus, we might have scaled even faster.

This invisible cost extends beyond business impact to your personal journey. Years spent grinding on a product with limited potential is time you'll never get back. The emotional and physical toll of dragging a boulder uphill—instead of riding a rocket—can't be measured in dollars alone.

Warning Signs Your "Working" Product Should Be Killed

How do you know if you're in this gray zone? Here are the warning signs I've observed across hundreds of companies:

1. Metrics That Plateau

At Stackify, our demo-to-close rate hovered around 30%. Pablo mentioned that truly exceptional products see 40%+ of demos convert to sales. If you're stuck below this threshold for months, it's a warning sign.

Another key metric: when you ask customers "How would you feel if this product no longer existed?" less than 40% say "very disappointed." As Pablo noted, this Sean Ellis test cuts through the politeness bias of NPS scores.

2. The Treadmill of Features

You keep thinking, "If we just add this one feature..." Yet each new capability only incrementally moves the needle.

Especially when you're the one that actually builds product... maybe this new feature, maybe this new thing. If I just add this one more thing, it's going to be the thing that takes off.

Honestly, it just rarely ever is.

This endless feature treadmill is exhausting—and expensive. I've built features that took months of development time that customers barely noticed.

3. Selling Is Always Hard

If you're constantly having to explain, educate, and convince people to buy your product, you likely haven't found product-market fit. True product-market fit feels like customers are pulling the product out of your hands.

At my first company VinSolutions, we had undeniable pull. We doubled revenue year-over-year and struggled to keep up with demand. Our biggest problem wasn't sales—it was scaling our implementation team fast enough!

Making The Hard Decision

So how do you actually make this decision? Here's a framework that combines Pablo's insights with my experience building and selling multiple companies:

1. Get Brutally Honest Feedback

Pablo shared how his team at Gymtrack thought they'd found product-market fit when gym owners enthusiastically praised their equipment usage analytics dashboard. But that enthusiasm didn't translate into sales.

The hard truth: customers will say nice things to your face, especially if they like you personally. What matters is whether they'll actually pay—and keep paying—for your solution.

2. Look At Top-of-Mind Problems

The Gymtrack story hit home for me. As Pablo explained, they built a product gym owners thought was cool but didn't solve their most pressing daily challenges.

Ask yourself: "Is my product solving a top-of-mind problem my customers wake up thinking about?" If not, you're building a nice-to-have, not a need-to-have.

3. Calculate Your Opportunity Cost

This is the hardest part. You must honestly assess: "If I started over with everything I've learned, could I build something with much stronger product-market fit?"

For every startup that limps along for years with mediocre traction, there's a founding team that could have created something extraordinary if they'd been willing to make the hard pivot.

"If people aren't pissed every single time your product doesn't work, then you're better off continuing to look for something."

The Courage To Start Over

I experienced this firsthand with my first company. We started with a solution for taking pictures of cars and uploading them to the internet, which seemed innovative in the early 2000s. But it wasn't until we pivoted to CRM software for car dealers that we found explosive growth—eventually selling to AutoTrader for $150 million.

That pivot wouldn't have happened if we'd stubbornly stuck with our original product vision.

As a CTO, product leader, or technical founder, you're uniquely positioned to recognize when a product isn't achieving its potential. You feel it in the constant struggle for growth, the lukewarm customer feedback, and the endless feature requests that never quite satisfy.

Having the courage to say "this isn't working" might be the most important decision you'll ever make.

As Pablo summarized perfectly: "An ounce of research is worth a pound of marketing or sales. Investing that time up front is so hard, especially when you have something that's kind of working, but it's totally worth it every time."

The question isn't whether your product is working at all—it's whether it's working well enough to be the best use of your limited time on this planet.

---

Join 58,914 others, follow me on LinkedIn.

Matt Watson is the host of Product Driven and co-founder of Full Scale, a global staffing company that helps businesses build and scale their engineering, finance, marketing, and admin teams. A three-time founder, he grew VinSolutions to $30M ARR before a $150M exit, later sold Stackify in 2021, and continues to share insights from his entrepreneurial journey through his podcast and this newsletter.