You are both right. That's why this is such a great story. It's a learning experience at every level!

My own take is that the poor guy's fundamental error was that he jumped the gun on outsourcing. He should have built the first hundred prototypes by hand on his kitchen table, with parts scrounged from one hobby store or another, then sold them himself. It sounds like he basically ended up doing that anyway, but if he'd done so in the first place he'd have invested less up front, gotten feedback faster, and gotten an intuitive sense of his market and the sort of product they'd prefer. Premature optimization!

The other, even more fundamental error was going into the novelty business in the first place.

It would've shown that his business idea was not economically viable, regardless of whether people wanted the keychains, before he had invested $10K in it. If there's no way a business will work, it's better to find that out quickly and cheaply than to invest more money in it.

This is harder than you would think. I took a year off from programming to help a friend market and sell actual physical stuff.

Often times a product that gets well received when you have a few samples made doesn't sell at all. Something you think kinda sucks and is just a 'filler' product in your line up turns out to be the most popular, and you run out of stock. Something sells well, you can get more made, so you order 500 more units, and then all of the sudden the product is not cool and nobody wants it anymore.

Oh, I know. My erstwhile startup seemed like a hot area when we began it a year and a half ago, and people in our target market said "Thank God" when I said what I was building. Now, nobody seems to care, and none of our competitors took off either.

I'm saying that even under his best assumptions about the market, his product was not profitable. He ended up getting customers after all, but even under generous volume and gross margins, it couldn't cover overhead.

Perhaps he should have tried sell the key chains for 5$.