So, I'm up a little late, picking through Mark Pilgrim's Dive Into Greasemonkey and cobbling together some simple scripts. (Here's a small peek into my tattered psyche.. I take great joy in following along with "Hello World"-type examples, but changing the customary greeting to "Hollow World.")
So what has me looking into all of this? The realization that Greasemonkey may have huge potential for lightweight prototyping for User Research purposes. (Did Lance suggest just such a thing to me a couple months back? Sometimes good ideas take awhile to really take hold themselves in my brain.)
It seems especially appropriate for those prototypes that just add elements or functionality to an existing site. If a prototype requires a bunch of new and novel interoperating features, then I'm thinking you'd be better off coding it up as a standalone entity.
But for that first type of prototype... the 'test out a new feature' kind, GM could be pretty cool. This type of prototype would really benefit from having an as-close-as-possible user experience to the real, live site or app. Which, of course, is what Greasemonkey is all about -- it allows you to insert arbitrary code into real, live, running pages.
I've got some potential prototyping work coming up that could be pretty focused. Just some small, discrete 'badgey-type' elements that would appear in a listing of search results. Maybe with some 'load a div on hover' kinda stuff, too.
In this instance, going "low-fi" (paper prototype, or static JPEGs of screen designs) could actually be more work: I'd have to recreate all of the elements that the existing app already possesses (and remember, we really just want to know what attitudes and acceptance of the new stuff is.) And, of course, it would be low-fi. So searches wouldn't 'really' work. And we'd have to do a lot of hand-holding and 'imagine if'-ing with the test users.
This would have a subsequent effect on the quality of the data we'd get back from them -- the more artifice and interruption we interject into the prototype, the more we run the risk of negatively impacting their perceptions of the stuff that we want information about. (I guess, put more simply, it would be better if the experience of using the prototype matched the experience of using the real application.)
It would criminally cool to let our test-users just interact with a live, running version of the site: perform searches that they select (and that are relevant to them); comment on an experience that already feels familiar to them; and get reactions mostly to the stuff that we've added via Greasemonkey.
Anyway, I'm sure there are folks out there already using Greasemonkey for similar purposes. Glancing at the Usability tag on userscripts.org (community repository for hundreds of helpful Greasemonkey scripts) is no help. (Thanks a lot folksonomy! What if folks' idea of Usability is different than mine?!)
Anyway, anyone trying this, or heard of any success stories?
Comments