I had a conversation with a friend recently who, after noting some of the recent Web 2.0 products, made essentially the "what were they thinking? The spent time/ money building that?" comment. He thought some market research could do some of these companies good. Then today I read this list of start-up rules on TechCrunch and noted number 7:
Don’t spend time on market research. Launch test versions as early as possible. Keep improving the product in the open.
SO which is it: to market research or not? For me, it all goes back to the user. If there is a user expectation about a product, where the new product is more of an incremental change, I actually think some market research can be beneficial. Think automobiles, clothing, etc. But for new markets, where users have no pre-conceived notions, market research adds little value. For examine, imagine trying to size the market or get focus group feedback on blogging tools in 1997. No one knew there was a market or knew what they wanted in a product.
The other consideration is time/ money. Doing market research, beyond anecdotal, is time consuming and costly. For web-products, where you can launch a site cheaply, and do some basic awareness marketing, you'll get feedback right away. If your nimble in your ability to react to the user community, you'll likely save yourself time and money. Other more resource intensive industries don't have this luxury.
Of course the real trick is understanding that many times people are averse to change and new things. As a result, outside an early adopter set, often times user feedback/ market research is negative. If you listen to them the result will be a product dumbed down to the wisdom of the crowds. That's what happens when people are left to lead themselves: they go to the place they already know.