Hah, I always love the reaction of people doing user testing for the first time. It's universally (as per the article) "I never really did this before and was AMAZED how people use the site vs. what I expected"
Some problems with the article.
1. Craigslist is fine but it will bias the users. If you ask people to do it for free, you're getting one subset of users (who are actually usually pretty decent). If you pay people, you open it up to a much wider, more motivated pool -- but you get "professional" test subjects.
2. Going to a random public place and asking people to try it is good for a shotgun approach, but if you're trying to get coverage for your target audience (particular if there is a certain demographic you need for monetization) this is a no-go.
As to the rest of the article's content... well, I don't want to rain on the guy's new-found usability evangalism, but this is pretty basic, heavily skewed stuff. It seems like he's passing on what his 'mentor' has told him completely uncritically. Particularly the Nielsen stuff. Ugh.
For people who do want to do free user testing for their site, I highly recommend ethnio.com. This allows you to target your existing site users and works great.
tim, totally good call. its insane to watch people. i remember at facebook people would click a certain link to do something and i yelled at them. why are you not doing this? i dont understand!!!
1. know what you mean about sample set. its hard to get exactly the right person but close enough is good for me.
2. for the public places, i have seen a few people on sports sites and go up to them. (we built a new sports site)
i've seen ethnio.com, good call. in terms of being an evangelist, im not at all. the real issue is that most startups NEVER do this, i wanted to document what i've done and if people want to use/benefit then good. if not, sucks for them.
I definitely appreciate the effort you've gone to. Unfortunately user testing without careful evaluation can wind up hindering or hurting a product. If you look at a site like 37signals (hah, had to make sure I got the capitalization correct - http://www.37signals.com/svn/posts/1088-how-not-to-apply-for...), they do a great job without it, and manage to upset the great Donald Norman while they're at it (http://www.jnd.org/dn.mss/why_is_37signals_so_1.html).
I do agree that startups should be doing more. You have done a good job of giving some steps to get started, so kudos.
you're right, the signals' products are usable, but they could definitely be better. i've yet to meet a website that couldn't be improved by watching customers and translating that feedback into an improved experience.
totally discounting the value of watching customers to improve a product means that the discounter has never had the visceral "i suck" feeling of watching a customer stumble through a product he's designed and built.
thanks tim! 37 signals is good but some of their stuff is not that great. look at basecamp. if you want to move a todo to another list. you have to delete it and re-enter. im not discounting what they do and their stuff is simple but not the end all. nice on the links.
I realize that it might be hard to break 522 days of habit, but consider adhering to standard English punctuation and capitalization in your comments. They will be easier to read that way.
Some problems with the article.
1. Craigslist is fine but it will bias the users. If you ask people to do it for free, you're getting one subset of users (who are actually usually pretty decent). If you pay people, you open it up to a much wider, more motivated pool -- but you get "professional" test subjects.
2. Going to a random public place and asking people to try it is good for a shotgun approach, but if you're trying to get coverage for your target audience (particular if there is a certain demographic you need for monetization) this is a no-go.
As to the rest of the article's content... well, I don't want to rain on the guy's new-found usability evangalism, but this is pretty basic, heavily skewed stuff. It seems like he's passing on what his 'mentor' has told him completely uncritically. Particularly the Nielsen stuff. Ugh.
For people who do want to do free user testing for their site, I highly recommend ethnio.com. This allows you to target your existing site users and works great.