Ask HN: What does a non-programmer bring to a 2-3 person startup team?

To answer your question, not all non-programmers are equal: much like not all programmers are equal. The types you see "seeking a technical cofounder" are the equivalent of 1990s "HTML programmers". On the other hand, a seasoned startup veteran who may not be actively coding (but, most probably holds a BS or MS in CS or EE) but has a track record of raising funding, selling the product, effectively handling business operations and the like has a lot to contribute (if you're at a stage where you need assistance in anything beyond product development: when this stage comes depends heavily on the kind of product you're building), even if his role won't be development.

I consider myself to fit in this category. I've always been a hobbyist programmer, but it is my damned penchant for humans over social code hosting platforms that has prevented me from becoming a professional developer, for better or worse.

Not necessarily. Offshore outsourcing of programming is quite cheap. The question is whether your MVP is so complex it warrants a $100/hour local programmer to handle that complexity. In many cases, it's not complex enough to warrant it, and paying $100/hour to a developer is inconsistent with building a MVP, because you're hiring a bells and whistles developer better suited to a large corporation with millions in profit at stake if the developer screws up. But, sure, if you're building something complex, you're shooting yourself in the foot by cheaply outsourcing.

"If you're willing to throw out the MVP code and have the new cofounder rewrite it, that might be something you could convince a developer to join up for -- but be prepared to "lose" the money for the initial MVP development."

That's exactly what I'm proposing. Who cares if you "lose" the money for the MVP development? If the MVP has shown traction, you haven't "lost" the money. You've paid cheaply for a traction proving MVP. That is money well spent, and now it's time to move onto the next level.

In any event, the idea that a developer would turn down an opportunity to be a founding equity member of a startup, which has a product that's already proven traction, because the MVP contained messy code is absurd. I'm not saying some developers wouldn't think that...I'm saying such developers are irrational.

"...but if you were looking for a partnership, I'd want to know what you'd spend your time doing."

Your attitude puzzles me. You have two hypothetical choices, if you were offered such a partnership.

1) Say no and continue to build software with a 90% chance of failure.

2) Say yes and build software with a 90% chance of success.

Even if the non technical cofounders sit on their asses and do absolutely nothing while you code all day, you'd be sitting on 33% equity in an already successful startup, when they took all the cash risk up front to prove the thing works. All you've got to do now is code and build on existing success. And you think that's an offer that you should scoff at?

In reality, the non technical cofounders wouldn't sit around and do nothing. This would be a group of people who've already proven they can go from zero to traction proving MVP on their own, and surely they'd continue to work in non-technical ways to make the product a success. I don't understand why you'd be so skeptical about what they'd do with their time, unless you think their success was dumb luck.

1. Finding customers 2. Building the product

Where finding the customers is something that they can be incredibly passionate about and actually affect. This is what the non-programmer worth their salt can add to a 2-3 person start-up. A skill-set aimed at focusing the pivots to those most likely to sell.

If you are selling big enterprise products priced above ~ $10K you absolutely need marketing/sales.

If you are lighting a powder keg, you got lucky and probably don't need much.

If you are trying to start a movement, you require hype, which can come from a dedicated community (e.g. tech) and may not require independent marketing. But you may need a community person. Or you may not be so lucky (e.g. your market is used car sales) and may at some point need a full-time SEO/partnership/etc. person.

If you are going "empire" you will eventually need these sorts of people, since they reach people that can't be reached otherwise. If you are going empire you will also need management people, since large organizations get unwieldy and there are lots of things they learn in b-school about wielding spreadsheets and org charts that will probably be useful.

Conclusion? I tend to think in general you don't need a marketing/sales person until you have a product, but it may be good to get them on board early in some circumstances, particularly if you are also searching for VC or trying to define the product better and need someone who can interface directly with potential clients and gather requirements, etc.