DESIGN, ART, GADGETS, FASHION, AND SAFEWAY

SUBSCRIBE VIA BLOGLOVIN bloglovin, IPAD, OR RSS

Wednesday, September 02, 2009

In Regards to Those Evil Focus Groups

| |

Focus groups are structured group interviews meant to reveal one thing - want. This isn't always explicit nor completely impossible to extract 100% solid info, but that's where judgment comes in. Knowing when to use focus groups is the first step.

You need focus groups when looking for desires, motivations, values, and first-hand experiences. In other words, attitudes and perceptions. Why do people value something over another?

Focus groups are not for getting usability information. They are not experts and will not outline why a certain interface fails in practice.

There are four types of focus groups: Exploratory (similar topic, observe), Feature prioritization (most attractive function), Competitive analysis (inclination to other products, why?), and Trend explanation (again, to answer, why?).

Now, how do we conduct a focus group? This is easy. All you need is a good schedule, a target audience, a scope (reasonable group size), and a carefully chosen set of topics. Your process should look something like this:

Target Audience:
Just make sure they are willing to talk about the topic at hand, and absolutely ensure that their opinions will not be judged (avoiding generational differences when speaking about music, boys and girls when speaking about cars, etc.). That is all you need.

Scheduling:
t - 2 weeks: Determine audience and scope, recruit. Select topics.
t - 1 week: Write a discussion guide, making sure it's engaging and thought out.
t: Conduct the groups and take great notes.
t + 3 days: After lots of reflection, start listening to the tapes.
t + 1 week: Analyze accordingly.

Recruiting:
Don't recruit those who know each other, those who know how focus groups work, and those who have wayyy more knowledge about the subject than all others in the group. You're looking for general balance, nothing more.

Then, Create a profile:
Demographic - Ages 20-55. Income irrelevant. Avoid $100k+.
Web Use - Has a PC. 1+ Internet experience. 5-10 hours a week for personal tasks such as shopping, comparing products, and info gathering.
Behavior - Has completed home improvement w/in 9-12 months. Cost $20K+.
Say, you can probably use this profile to recruit a couple of people. Most books say get 8-12 people, but for user experience, we're looking at 6-8 at most. For your Scope, you'll have to work systematically, using the first few groups to establish a trend and the later few as confirmation. Never use just one group to establish a conclusion.

When talking about Topics, ensure comfort by wording effectively. A Guide will help as well. Carefully ordered, nondirected, and open-ended questions will provide you with specifics and lend you personal and unambiguous experience data that will be priceless.

During the actual discussion, lay down ground rules, make everyone comfortable by explicitly dispelling any curiosities that may arise, and start sending out those Probes (aka questions), but strategically order them to avoid bias towards any subject. End it with a smile and ask them for tips to improve the session.

Now, let's Analyze. Collect transcripts (written by a helper, hopefully), quotes, observer opinions, and tapes. Create a quick hypothesis and start coding (short names categorizing and describing trends) the comments. Data Extraction time. We've got mental models, values, stories, problems, and competitive analysis. Play these into your hypothesis, carefully gauging the magnitude of trends, creating lists of components that play into the decision making, and toss in quotes for good measure. Fin.

EXTRAS: Always record the session's audio. If possible, throw a hidden camera in there to capture expressions and body language. And as a moderator, you must also be able to vanquish all that nasty "group-think." It's bound to happen, so be prepared to dispel it.

0 comments:

Post a Comment

Popular Posts


Archives