Bring a laptop that you don't care about. Drunk people spill.
The $5 Guerrilla User Test
11–20 of 49 posts
Re: The $5 Guerrilla User Test
#12I've done the sober equivalent to this (coffee shops). Not only did it point out some usability issues, like, "oh, he thinks THAT'S a button," it also proved to be a great way to generate ideas. Most of the input we received was dead wrong, but we had to think about why it was wrong, and that got us thinking in ways we hadn't thought before, which led to good ideas. I suspect the drunken version would be more fun and just as useful, assuming you are producing a consumer product. As for B2B, if your domain is specific, this might not be the most valuable use of your time.
Re: The $5 Guerrilla User Test
#13What other types of guerrilla testing have people tried before? This sounds like a fairly witty idea for gathering user interaction information, but I wonder if it's actually functional.
I believe there's at least one website review service that is build on top of Mechanical Turk, if that helps any.
Re: The $5 Guerrilla User Test
#14I don't see myself doing this.
Re: The $5 Guerrilla User Test
#15One additional note that should be added: results will only be useful if your desired customer base is represented well by people at a bar.
FTA: "Bars are wonderful at segmenting by demographic. Match the bar you’re going to with the user population you want to target. Different bars will produce slightly different results but the variation is not huge."
Re: The $5 Guerrilla User Test
#16So you want programmers to talk to people in real life. Interesting.
Re: The $5 Guerrilla User Test
#17I experienced the same effectiveness (95%-99%) of getting the "guy/gal reading the newspaper" to volunteer for the study first hand. Although, this was more focused in cafe environments.
Our team was trying to get beta users signed up for a location-based iphone app in the Boston/Cambridge area and most people reading newspapers in coffee shops or cafes are frequently willing participants. I even noticed that 7/10 times these people will engage with you in conversation regarding feedback/UX improvements etc.
In parallel with Rodyancy's experience, some of the user feedback was arbitrary (one guy started comparing our application to time/space continuum theories...etc) but nonetheless, I would advocate approaching the newspaper guy/gal as they are already on the knowledge quest and usually are more open-minded.
Oh, and avoid the book/textbook readers. These people are either really engaged in novel (imagine tapping on someone's shoulder during a movie) or are stressed-out students.
Re: The $5 Guerrilla User Test
#18What other types of guerrilla testing have people tried before? This sounds like a fairly witty idea for gathering user interaction information, but I wonder if it's actually functional.
I believe there's at least one website review service that is build on top of Mechanical Turk, if that helps any.
Thanks, that's good to know. I know someone once mentioned using Mechanical Turk and that those sometimes ended up being their most loyal users, but I cannot recall the context. Has anyone used these services and what results did you see?
Re: The $5 Guerrilla User Test
#19One additional note that should be added: results will only be useful if your desired customer base is represented well by people at a bar.
FTA: "Bars are wonderful at segmenting by demographic. Match the bar you’re going to with the user population you want to target. Different bars will produce slightly different results but the variation is not huge."
Yeah, I don't know a bar frequented by people interested in the Semantic Web or web-crawling.
Re: The $5 Guerrilla User Test
#20Earlier quoted context omitted.
I believe there's at least one website review service that is build on top of Mechanical Turk, if that helps any.
Thanks, that's good to know. I know someone once mentioned using Mechanical Turk and that those sometimes ended up being their most loyal users, but I cannot recall the context. Has anyone used these services and what results did you see?
kartme posted a good post about changes they made with usability testing. they mentioned uservoice which is on turk if I remember, they mentioned doing a followup post explaining more about what the did during testing.