A Diamond in the Shit

We have countless conversations in tech around design, user experience, branding, performance, frameworks. We call these things important. They feel important. We build entire careers around them.

But they're not.

There is only one thing that matters: value. Everything else is a crutch for avoiding it until there is actual proof of value.

A big diamond ring is a valuable product. Most products are not big diamond rings - they're void of any defensible value.

Rather than confront that, teams invest in everything except creating value. Better design, better messaging, better performance - believing that all that's missing is the perfect packaging.

But put a big diamond ring at the bottom of a pile of shit and you'll see how little packaging really matters. Once people know it's there, all you have to do is sit back and watch them roll up their sleeves.

I was reminded of this recently. Our entire company moved to Claude Code in the terminal - technical and non-technical people alike. Not because it was a better experience. Not because the interface was pleasant. Because it delivers the most value. Full stop.

This is the same story as Craigslist, AWS, Costco and many others. Products that deliver so much value that friction, experience and form factor were proven irrelevant. Companies that never got distracted by bells and whistles - bells and whistles that only become necessary when you fundamentally have a product no one would reach through the pile of shit for.

Most teams don't consciously ignore value in favor of polish. They conflate optimization with value creation and genuinely can't tell the difference. They believe polish is the product.

This happens because optimization is legible. You can measure conversion rate improvements. Track NPS. A/B test button colors. 

However, value creation is murky. It requires asking "what would make someone's life dramatically different?" and sitting with the discomfort of not knowing, of being wrong, of building something that might not work.

So teams default to what's measurable. They optimize the shape of the shovel instead of asking whether anyone wants to dig a hole.

The leverage gap between optimization and value creation is categorical. Optimization captures existing value. Creation generates new value. One is linear improvement on a fixed curve. The other shifts which curve you're on entirely.

Before you optimize anything, you need a clear-eyed answer to one question: if we stripped away every optimization and delivered this capability through the worst possible interface, would anyone still want it badly enough to fight for it?

If not, you're polishing.