The client is wrong. Ship it anyway.

A client once asked me to break his own website.

He didn’t know that’s what he was asking. But that’s what it was.

Here’s the story — and the one skill it taught me that no technical course ever will.

The setup

I was building a website for a Korean manufacturer entering the US market.

The site had one job: make American investors and partners take the company seriously.

English content. US messaging. That was the whole point.

Then the CEO asked me to make the .com serve Korean.

We already owned the .co.kr domain. The Korean site had a home. Putting Korean on the .com would confuse the exact audience we built the site for.

I knew it was a bad idea the second he said it.

But here’s the thing

He wasn’t being stupid.

His revenue came from Korea. His phone rang with Korean numbers. And every time he typed his own company’s .com — the domain with his name on it, the one he paid for — he saw a website his customers couldn’t read.

From his chair, I was the one with the dumb idea.

You have 3 options

When a client asks for something wrong, there are three moves:

1. Fight. Bring the data. Cite best practices. Win the argument. And lose the relationship. You just told a founder he doesn’t understand his own website. He’ll remember that in every conversation you have after.

2. Fold. Do what he asked. Break the strategy quietly. Let the project fail slowly enough that nobody can blame any single decision. This is what most people do. It feels agreeable. It’s abandonment with good manners.

3. Find the third door. Ignore what each side is asking for. Find what each side actually needs. Then check if those needs really conflict.

They almost never do.

The third door

What did the CEO need? To type his .com and see his business, in his language.

What did the US strategy need? Americans landing on that same domain and seeing an English site built for them.

Written that way — no conflict. Just two audiences.

So I set up geo-based routing.

Visitors from Korea → Korean site. Visitors from anywhere else → English site. Same domain.

The CEO opens the .com in Korea and sees exactly what he wanted. A reviewer in the US opens the same URL and sees exactly what the strategy required.

Nobody compromised. Because there was never a conflict — just two people looking at one URL from two different chairs.

Is it elegant? No.

Purists will mention VPNs. Search engines. The “right” way to build multilingual sites.

They’re not wrong. Technically.

But the technically-right answer that kills your relationship with the founder isn’t the right answer. It’s just the one that lets you feel smart while the project dies.

What this is actually about

I’ve spent my career where the plan on paper and the reality in the room never match. The military. A global NGO. A fast-growing startup. Now client work.

The pattern never changes:

Projects don’t get killed by technology. They get killed by the fight over who’s right.

The client is sometimes wrong about the how. He’s almost never wrong about what he needs.

Build for the second thing.