Monday, March 16th at 05:00 PM
Presenters:
Mark Trammell – Digg
Juliette Melton – Lumos Labs
Nate Bolt – Bolt|Peters
Carla Borsoi – Ask.com
Andy Budd – Clearleft Ltd
Useful Tools:
- User testing – more of a design tool and not a research tool
- Interviewing users – direct questioning, or interviews
- Talking directly to users
- Analytics and specific web testing – loyalty of users coming back to site – the end result
- Customer service feedback – satisfaction level with a product
- Behavior vs. Opinions
- One on one testing – simply watching what faces people make
- Surveys – a good way to access a lot of voices at one time, and the ability to break down demographics or pull out specific quotes.
Useless Tools:
- Eye tracking "seems like science, but not" – BS! – Why aren’t people clinking on this link? You don’t need eye tracking to tell you it is small or out of the way.
- One on one interview is much more useful than what you can get out of a focus group. A brainstorming session with strangers motivated by candy, One-on-one is better. You CAN get useful info out of a focus group if done well.
Concerning yourself with user research? Steve Jobs: "I don’t"
It all comes down to genius design (Ives at Apple or 37 Signals). "We don’t need to do research!" These people know exactly who they’re designing for, themselves, or Steve Jobs. If people (like them) find it useful, then great. We all feel like we have a better insight on our users than we actually do. Apple is actually doing research in the background, they’re building prototype after prototype and testing internally, so that testing IS happening, just not in the public. If you define research as any sort of internal testing, it counts. Flickr started as a game site and turned into an incredible photo site based on what users wanted out of it.
Remote Testing:
Observing someone through some sort of remote test (from another location). Lab testing is BS. "Imagine you’re buying something at this soulless computer in a white room with a 2-way mirror."
How do you poll users?
Silverback is a nice app to record the user as they are interacting with your website. It captures video of their facial expressions.
How do you pull in the things that are wrong into the research? Research can be boring – getting people involved can be easier to make it more exciting (food/drink?!?). There is sometimes an empathy with a user who is struggling.
Sharing the feedback of the test, or the decisions made off the results.
If you have to bribe your designers/developers to care about their users, they are not good for your team. They should WANT to be involved.
Ah Ha Moments
Starting with paper prototypes, we learned that it was not the right way to go. Getting it early was helpful.
Even though customers trusted the company, but not a certain product over a competitor. Hearing it actually come from users, with evidence, it make it around and created change.
Quantitative vs. Qualitative Research
Quantitative: A/B testing on a group statistically important (inside)
Qualitative: "This one user said" gives the story behind it. (ammunition)
How do I pay for this?
You have users, just find some people and ask. Some of the best research is cheap/free research.
Silverback
Surveymonkey
Agile Testing
MS put this together for Age of Empires game. They schedule time between tests to actually fix the issues between the tests, and over a few days, you get things fixed from group to group. Right Methodology
Biggest Mistakes Made
- Testing for less time in the UK vs US
- Biggest failed study with kids from Chile/Mexico none of them speak english
- Getting people you know for user testing (don’t fit the people who really use the site).
- Only drawn bloog once in a usability test (oops)