Guesses vs. Data as Basis for Design Recommendations (Jakob Nielsen's Alertbox)

I am always telling people to talk to their customers. Find out what they're doing, get data about them, don't just make assumptions. Well, here's some empirical data that shows that this is a good plan: Guesses vs. Data as Basis for Design Recommendations (Jakob Nielsen's Alertbox). In a nutshell, Nielsen shows that people who relied on even very small samples of data (observing 1 or 2 people) were more likely to arrive at correct assumptions around design functionality than those people who simply relied on their "gut feeling" or their own personal experiences. What does this mean to you? It means that if you're making a change to your design and you can show it to even just one of your customers ahead of time, you'll make better decisions about how that design will affect your customers. Testing your pages is critical, but you need to test more than just browsers. Find a friend, colleague, relative, or co-worker who is in the demographic for your site and have them look at your redesign. If you're trying to make the site easier to use, and it's even harder for them, you should return to the drawing board. Yes, they aren't a representative sample, but even one person is better than zero.