I don't know if I'll be downvoted for this, but - no. When I design, I build it for one person and one person alone - myself. I add the features that I want to have access to. I make the layout one that I find attractive and usable. I consider myself to be my product's first user. While I can see market research helping out a lot, I've always felt that the smartest thing to do is to make it for yourself and take feedback only after you're entirely satisfied with it.
It's a process that I think helps a lot if you're working with a really original concept. The user doesn't always know best. If you're doing it right, chances are you know exactly what you want, and that is something that you ought to follow first and foremost.
When I first conceptualized my site, I took the idea and told writers (the target audience) all about it. Most of the time, they couldn't visualize it whatsoever. When I asked them what DIDN'T work in sites today, they would say either that they had no clue, that there WAS no good solution, or that things were satisfactory as they were.
Last month, I demoed the concept HTML page to those same people for the first time, and you could absolutely feel the minds shifting. People abruptly understood the concept behind the page. The design of the page worked perfectly for them as it had been originally conceived.
I understand doing market research if you're going out and just stumbling upon an idea, looking for something that will work. But if you've got an idea, something that's never been done, then you take it and you push it and you wait until after it's done to see how people react to it. The result is a much more honest and effective site that way.
I'm sure other people disagree with me, so feel free to debate this point back at me.