Several articles came out since Joel Spolsky's OnStartups answer about splitting shares, some very cogently arguing against a 50/50 split, and proposing alternative methods. I've waited until a few accumulated before posting a summary.
50/50 isn't a business decision, it's a compromise
It goes on to propose a relatively complex system to get this hard decision done up-front and avoid later arguments.
50/50 partnerships can be hugely unstable - even if you've been friends since high school.
Most senior employees who join are given 2% if they join early. Maybe they get up to 10% if they joined REALLY early and were senior. Who gets 30%? Nobody. That's who. So trust me when I tell you that you can hire incredibly talented people for 30% of your company. Or 20%. Let's be honest - even 10%.
It's worth noting that both Mark and Dan had a 50/50 split in their first startup (they both say so in their articles). This means they know first-hand the problems it can cause.
And I second with emphasis the focus on fairness. Founding teams that allocate the founders equity fairly stay together a lot more than founding teams where one founder has a much better deal than the others. The same is true of venture capital firms. The most stable venture partnerships are those where the partners share in the carry equally or near equally. At the end of the day, this is as much about respect as it is about money. And when people feel disrespected, they are going to leave at some point.
So, what to make of these two apparent contradictions, both of them argued very convincingly, by people with vast amounts of experience in the domain?
Well, usually when you get this kind of situation, where both sides are convincing, both sides are credible, both sides make sense, that points to a very interesting "hidden variable", a disagreement that's not obvious, and usually hidden in the meaning of commonly accepted words.
In this case, I believe the hidden variable is inside the meaning of the word "friend".
Fairness and self-interest
There are many different types of possible relationships between people. To take just two, some people form relationships around fairness, and others form them around self-interest (and most use both with different people).
There's nothing wrong with either of those. Both occur frequently. Typically, older relationships tend to be based around fairness, but not always. And it's not that rare for one person to think that the relationship is based around fairness, while the other person is focused on self-interest. This is not a bad reflection on either friend, just a reality.
If you structure a company's shares based on fairness, but your relationship is based on self-interest, you're headed for the trouble Mark and Dan mention, as soon as you hit a hard decision. If you structure a company's shares based on self-interest, when your relationship with your cofounder is based around fairness, then the minute you use the power which you gave yourself to overrule the other person's decision, you've probably lost a cofounder and gained a competent but demotivated employee. And you'll have to deal with the demotivational aspects of the lack of fairness of the split forever.
My view, then, is that you should base your equity split on the type of relationship you have with your cofounder. There is no "one size fits all" solution. Make sure you fully understand what the dynamics of the relationship are before you decide on a split, that you know where each party is coming from, what they want, what they look for in this friendship. Then make sure that split extends and strengthens your relationship rather than undermine it.
And don't forget to read this, especially the comments, for a view of what happens when it all goes wrong.
If you read this far, you should follow me on twitter here.