When Doubling Down Is the Worst Bet You Can Make

There's something deeply satisfying about optimizing a system. Tweaking the process, shaving off inefficiencies, watching the numbers nudge upward — it feels like momentum. It is momentum, technically. The problem is that momentum doesn't care about direction. You can be moving very, very fast toward a dead end.

This is what we might call the momentum trap: the psychological and practical phenomenon where the energy you've already invested in a system becomes the very thing that keeps you from escaping it.

The Sunk Cost in Disguise

Most people have heard of the sunk cost fallacy — the tendency to keep pouring resources into a failing endeavor because of what you've already spent. But the momentum trap is sneakier. It doesn't feel like desperation. It feels like diligence.

You're not staying out of desperation. You're staying because the system almost works. Because you've gotten better at running it. Because leaving now would mean admitting that all those optimizations were applied to the wrong foundation.

Cognitive scientists have a term for this: escalation of commitment. The more you've refined something, the harder it becomes to evaluate it objectively. Your brain starts defending the investment rather than assessing the outcome.

And here's where it gets practical: in careers, in businesses, in creative projects, this pattern kills more potential than outright failure ever does. Outright failure is obvious. The momentum trap is quiet.

The Startup That Learned to Quit

Consider the now-famous pivot of Stewart Butterfield's company Glitch — a quirky online multiplayer game that never quite found its audience despite years of refinement. Rather than continuing to optimize a product that wasn't resonating, Butterfield and his team made a radical call: abandon the game entirely and build something new from the internal communication tool they'd developed to run their own operations.

That tool became Slack. At the time of its acquisition by Salesforce, it was valued at over $27 billion.

The lesson isn't that you should always quit. It's that Butterfield and his team were honest enough to ask a question most founders never get around to: Is this system worth improving, or is the system itself the problem?

The Three Signs You're Polishing a Sinking Ship

So how do you know when you're in the trap? There are three signals worth watching for.

The improvements are getting smaller. When your gains shrink with each iteration — when you're celebrating 2% efficiency gains after years of effort — you may be approaching the ceiling of a flawed model. True progress compounds. Plateau-level gains often signal a structural problem, not an execution one.

You're solving the same problem repeatedly. If a particular friction point keeps resurfacing no matter how many times you address it, that's not bad luck. That's your system telling you it was designed around a faulty assumption. Patching the same leak in a boat doesn't mean the boat is seaworthy.

The energy cost is disproportionate to the output. When a system requires heroic effort just to maintain — not to grow, just to hold position — it's often because you're fighting the natural grain of the architecture. Good systems get easier over time. Broken ones demand more.

Starting Over Isn't Failure — It's Information

There's a cultural bias in American professional life that treats persistence as inherently virtuous. Grit is celebrated. Pivoting is tolerated. Starting over is quietly viewed as a kind of defeat.

But some of the most durable careers and companies are built on the wreckage of prior attempts. Instagram started as a check-in app called Burbn. YouTube's founders originally envisioned a video dating site. The history of innovation is littered with smart people who had the courage to recognize a bad foundation early — and the self-awareness to walk away before sunk costs became identity.

Psychologist Carol Dweck's research on growth mindset is often cited in the context of resilience, but there's an underappreciated corollary: a genuine growth mindset doesn't just mean you persist through failure. It means you're capable of accurately diagnosing what kind of failure you're experiencing — and knowing that some failures are signals to change direction entirely, not double down.

The Practical Demolition Test

Here's a framework worth keeping in your back pocket. When you're deep in optimization mode on any system — a business model, a job search strategy, a creative workflow — ask yourself three questions:

1. If I were starting from scratch today, knowing what I know now, would I build this same thing? If the honest answer is no, that's worth sitting with.

2. Am I improving performance or managing decline? There's a difference between getting better and merely slowing down the deterioration. Be honest about which one is actually happening.

3. What would I have to believe for this to work? Sometimes articulating the assumptions your current approach rests on makes it obvious that those assumptions were never really valid — you just never examined them out loud.

None of this is an argument for impulsiveness. Starting over has real costs. It takes time, capital, and emotional energy. The goal isn't to abandon things at the first sign of difficulty. The goal is to distinguish between a system that needs more refinement and a system that is fundamentally broken.

The Fastest Route Forward Is Sometimes Backward

Counter-intuitively, the people who reach their goals fastest are often not the ones who never quit. They're the ones who quit the right things at the right time. They don't confuse momentum with direction. They don't mistake activity for progress.

If you've been grinding away at something that isn't quite working — optimizing, adjusting, iterating — it might be worth stepping back long enough to ask whether the problem is your execution or your foundation.

Sometimes the bravest, most strategic thing you can do is light a match.