Personas vs designing for yourself

Since Alan Cooper wrote about the idea of "Personas" – fictional people designed to represent a product's most important users - the idea has become a core method in the field of "interaction design". But recently a few notable designers have created some controversy by advocating a different approach: designing for your self. A while back, Jason Fried of 37Signals fame gave his opinion on the subject: "We don't use personas. We use ourselves." Recently Joshua Porter of Bokardo wrote a great post on the topic. He recognized the usefulness of personas but he argues that they aren't always needed and, in fact, it is best when you can design something without them. It is best when you can design something that you will use yourself. Not only do you know what you need but you will be passionate to build it. I agree.

But what I find interesting about Jason Fried's post is that he suggests eliminating personas altogether: "So if you can’t design something for yourself, design something for someone you know. Get that person or people involved in your project early on. Basing your decisions on a matrix of personality traits isn’t what I’d recommend if you really want to build a great product." I think it's clear that if you can work on a project that you will use yourself, design it for yourself. But I think the question is, what if you run into a project that you aren't passionate about (like, say, a medical application). Do you use personas? I'm not sure. If you aren't passionate about the medical field, do you make a persona based on user research? Or do you get the surgeon involved, ask him questions, and design it for him?

In designing Homeschooling Together, I've toyed with the idea of personas, but I never used them. For this particular project I've adopted Jason's approach: I'm designing it for my mom. We'll see how it goes when we release the beta but so far we've gotten fairly positive responses from other homeschooling moms.

UPDATE: A few more thoughts. My experience is quite limited so I can't take a solid stance on the issue but I don't think personas should be thrown out altogether. I was looking through the comments on Jason's post and it is evident that many people (such as Cooper) have used personas with great success. Based on this, I think personas can be a powerful tool when used correctly (though I've never used them myself). But I also think there is a good argument for using real people instead. In short, my current opinion on it is, if your designing something that your not passionate about and you think you can use personas effectively, go for it. But designing for yourself is probably the most effective, so do it whenever possible.

UPDATE 2: I've been doing research and came across an interesting response to Jason's post by Terry Bleizeffer. Terry argues that personas are not about getting rid of talking to people, they are the output of talking to people. You can't have the target user with you 24/7, so personas are a way to understand the target user without him being on hand (and they are the result of talking to actual people).