Chalkmark aims to address the “challenge of user testing when there is minimal budget or time” – a worthy goal heading into cash strapped 2009. It’s made by Optimal Usability, a New Zealand based company who also make OptimalSort. Chalkmark is currently free and in still Beta, though the site blurb implies that it will be a freemium service once it takes off.
How does Chalkmark work?
- You upload a series of still images as the stimulus for your study.
- These can be scanned paper prototypes, wireframes, or anything else you can convert into a still image.
- For each one, you are able to ask a question, of the form “Where would you click to do X?”.
- When the user clicks on their desired location, the coordinates of the click are saved, and the next question/image pair is loaded.
- The test is given a URL which you can distribute any way you please for recruitment purposes (it would probably work quite well in tandem with Ethnio).
- The output is a heatmap for each question, showing where the users clicked.
Chalkmark’s current limitations:
- It seems to be still in Beta (I’ve experienced a few bugs, but I’m sure they’ll be ironed out soon).
- There’s no way of screening, profiling or segmenting you participants.
- You can’t add any other kinds of questions (there’s no comment or questionnaire component).
- You’re limited to still images, with one question per image. You can’t add branching logic, so interactive prototypes aren’t possible.
This clearly isn’t a one-size-fits-all research tool, but does look like a clean, lean and well designed web app that I’m sure would be very helpful in certain situations. Also coming soon from Optimal Usability is TreeJack, a reverse cardsorting tool that helps you validate large IAs by giving tasks to real end users. I have to admit, though, personally I’m holding out for Webnographer from Fera Labs to be launched so I can start doing some full fat remote usability testing in the new year.
Hi Harry,
Just read your post about ChalkMark and I think it looks pretty cool. Wondered what you thought about using Amazon Mechanical Turk to recruit a some testers?
Sorry, I meant to test say 100 users albeit unsegmented with no profiling…? Is more the merrier?
Hi Phil,
the thought had occurred to me. There’s no reason why you can’t add a questionnaire to the task and gather some profiling information.
It would definitely be fun, quick and fairly cheap – but I would worry about the match between the turkers and your real user-base.
Hi Harry,
Thanks for the write up about Chalkmark. Your observations of its limitations are very astute! We definitely have all of those constraints under current consideration. In fact, we hope to be implementing some of those new features very soon.
As for recruitment, we currently envisage most users will be doing adhoc testing by using friends, co-workers, family etc for some informal testing. But for representative end users, yes, we definitely Ethnio. We’ve been talking to them about how we can perhaps work more closely together to make that recruitment process more invisible.
Any feedback at all, please, love to hear it both good and bad.
Cheers from a sunny Wellington morning.
Sam
Optimal Workshop