There's a great article by Robert Strauss in Stanford magazine describing how his startup failed.

Kudos to him for writing the article.  It's a golden opportunity for us to highlight some of the things he did wrong.  

I'd love to hear your comments -- if you decided to go into the exact same business he did (selling a condom on a keychain), what would you have done differently?

The most glaring immediate error I see is that he pursued large quantity orders without fully testing the market.  He was so focused on buying units in quantities of 10,000 that he didn't gauge demand and define success by making just a few prototype units first.

He could have done a lot with just a few dozen sample units, especially iterating on the design.  At one point after ordering 10,000 units, he realized that the market wanted a different type of product with a condom that was actually usable.  This could've been easily avoided by creating just a few test units first.

Also keep in mind that this article was written in 1999.  Selling over the web would've given him a huge new distribution channel that didn't exist back then.  So there's also a lot of room for innovation around things like:

• Ghetto testing using AdWords to get sales before you even create the product
• Selling on sites like Etsy
• Using tools like Helium to accept credit cards right on your site

What else would you have done differently?  I'd love to hear about it in the comments.