from Observing The User Experience - A Practitioners Guide To User Research by Mike Kuniavsky


Focus groups are structured group interviews that quickly and
inexpensively reveal a target audience's desires, experiences, and
priorities. Sometimes vilified by their association with dishonest
marketing, they do not deserve the notoriety with which they are often
treated. They are neither the panacea for curing bad products nor
pseudoscientific voodoo used to justify irrational decision making.
When guided by a good moderator, carefully analyzed, and appropriately
presented, they are an excellent technique for uncovering what people
think about a given topic and, especially, how they think about
it. They reveal what people perceive to be their needs, which is
crucial when determining what should be part of an experience and how
it should be presented.

Originally called "focused interviews," focus groups were developed
as a social research method in the 1930s, then refined as a method for
improving soldiers' lives during World War II and embraced by marketing
in the 1950s. As such, they're probably one of the oldest techniques
for researching the user experience. A focus group series is a sequence
of tightly moderated group discussions among people taken from a thin
slice of a product's target audience. The discussions are designed to
make people feel comfortable revealing their thoughts and feelings.

In software or Web site development, focus groups are used early in
the development cycle, when generating ideas, prioritizing features,
and understanding the needs of the target audience are paramount. They
can tell you what features people value most highly and why they
value them that way. As a competitive research tool, they can uncover
what people most value about competitors' products or services and
where those products and services fail. Sometimes they even reveal
entirely new competitors or applications for the product or service.

By providing a way to hear a lot of firsthand experience in a short
time, they can give development teams an early, solid foundation from
which to analyze the product and its users' needs. And as a watchable,
tangible, jargon-free method, they engage members of the company in
product development who would not normally have the opportunity,
expertise, or time to participate in the user experience research

In short, they provide a unique opportunity to see reality from the
perspective of the user quickly, cheaply, and (with careful
preparation) easily. Two people working part-time can set up a series
of focus groups, run them, and analyze the results in three weeks.
Conducting contextual inquiry research with a similar number of
participants, however, can take a full-time researcher almost twice as
long. A survey, furthermore, would need to sample a much larger number
of people—significantly increasing the complexity of the logistics and
analysis—while providing less understanding about the motivations and
attitudes of the respondents.

That's not to say that focus groups provide the same information as
contextual inquiry or surveys. All three techniques uncover different
and equally useful information—but focus groups can be an inexpensive
route when you need a lot of solid information in a short time.


When Focus Groups Are Appropriate

The purpose of focus groups is not to infer, but to understand,
not to generalize but to determine a range, not to make statements
about the population but to provide insights about how people perceive
a situation.

—Richard A. Krueger, Focus Groups, p. 87

Knowing when to use focus groups is one of the keys to using them
successfully. Although the technique is straightforward and flexible,
it's not applicable to all cases or at all stages in the development of
a product.

What Focus Groups Are Good For

Focus groups are good at finding desires, motivations, values, and first-hand experiences;
in other words, they're a tool to get at people's attitudes and
perceptions. A focus group is an environment where people (ideally)
feel comfortable revealing their thoughts and feelings. This allows
them to share their view of the issues and assumptions that lie at the
core of an experience and to relate them to real-world situations.

When coupled with contextual inquiry and task analysis, a complete picture of how and why people are behaving right now can be built, before a single line of code is written or a single screen layout is sketched on a whiteboard.

As tools for competitive analysis, focus groups come in early in
product development, though they can also be done during redesign or
update cycles. Understanding fundamental issues and perceptions is
generally most needed near the beginning of a product development
cycle. That's also when the development team is trying to nail down
what problems their product is supposed to solve, how it's supposed to
solve them, and why it's valuable for consumers to use their solutions
versus all others. Likewise, by bringing users of competitive products
in early in the process, it's possible to find out why people value the
competition, what they feel are the most critical features, what
regularly bothers them, and where they feel the competitors fail. Apart
from the obvious marketing angle, this information can immediately
drive feature and interaction development, defining a user experience
that's closer to what the target audience wants before resources have
been committed.

Later in the development cycle, the technique can help identify and
prioritize features. Knowing why people value certain features can help
determine what gets developed and in what order. Moreover, since focus
groups can act as brainstorming sessions, it's possible to achieve a
synergy in which participants generate more ideas together than they
could have come up with on their own.

One example from my practice was a site for Web novices that
concentrated on news and information about developments in the Web
world—which companies were doing what, what software was coming out
when, and so forth. In conducting a series of focus groups with their
target audience, we discovered that most of the people who were new to
the Web were also new to computers in general. They
weren't interested in Web industry developments nearly as much as they
were interested in knowing the basics of how their software worked and
where to turn for help. Moreover, since the process of getting on the
Web was difficult for nearly all of them, they were tired of dealing
with it and uninterested in installing or learning new software. They
wanted to know what they could do with it and how they could use it to
help the rest of their lives. Based on this information, the site
decided to de-emphasize its Web news and information and to emphasize
the self-help nature of the Web, collecting information about popular
software products in one place.


Focus groups uncover people's perceptions about their needs and their values. This does not mean that they uncover what people actually need or what really is valuable to them (in some objective sense); it's just what they think they need and what they claim
is valuable. However, relativism (and italics) aside, knowing
perceptions of needs is as important as knowing the needs themselves.
Ultimately, people are going to judge a product based on what they
think it does and how well that matches what they think they need. It
doesn't replace accurate functionality or needs assessment, but the
closer a product's presentation of its services matches people's
perceptions of their needs, the more they're likely to use it.

What Focus Groups Are Not Good For

First and foremost, focus groups are not a way to get usability
information. There's no good way for a group of people to tell you
whether they will be able to use a certain interface or a certain
feature. They can tell you whether they like the idea of it, but they
can't show you whether they can use it in practice.

Second, focus group results are impossible to numerically generalize
to a larger population, so they can't replace surveys. Specifically,
focus groups create generalized models based on observations by an
analyst. These models are assumed to apply to groups similar to the
ones interviewed, but because they are not statistically significant
samples, there's no guarantee that the proportion of responses in the
group matches that of the larger population of users. This is an
important distinction between focus groups and surveys. Because surveys
are statistically significant (or should be), then the proportions
observed in them can be extrapolated to larger populations. With focus
groups, there is no such assurance.

Thus, although it's an inadequate technique for investigating the
prevalence of a phenomenon in the entire target audience, focus groups
can give you a really good idea of why the audience behaves how
it does. Once the "why" has been determined, it can be verified through
statistically significant research, such as a survey. However,
identifying trends is often enough to act upon, making specific
quantification unnecessary. Just as it can be sufficient to know the
block a business is on without knowing the specific address, it's often enough to know that a phenomenon is widespread without knowing its exact magnitude.

Focus groups can create situations that are deceptive both to the
participants and to analysts who literally interpret statements made in
focus groups rather than extracting their underlying attitudes. This
can be seen in the plethora of products (such as many feature films)
that are made worse, rather than better, by misinterpreted focus group
results. An amusing, if somewhat apocryphal, example of this comes from
comic books: in an attempt to give Superman fans what they wanted, a
focus group of comics consumers (10- to 12-year-old boys) was asked
what kinds of figures they admired. Their replies were interpreted
literally, and for a while in the 1960s, Superman did whatever the
focus groups decided, leading to a string of surreal stories of the Man
of Steel working as a police chief, dressing up as an Indian, or
meeting George Washington (and to Jimmy Olsen, a meek supporting
character, turning into a giant space turtle). It led to a kind of
creative bankruptcy and an impossibly convoluted storyline that had to
be eventually scrapped entirely, the comic starting over as if none of
those stories had happened.

Finally, focus groups are not useful in situations where it's
important to prove a point or to justify a position. The data
collection process and analysis is based on multiple levels of human
judgment. As such, the results can be called into question when used as
a basis for proof, and rightly so. This is an excellent exploratory
procedure; it produces deep insight into motivations and thinking
processes, but it does not (and cannot) be used to unequivocally prove
or disprove anything.

Four Types of Focus Groups

There are four common types of focus groups in software or Web
development. The type of group you choose depends on the types of
questions you want to answer, which in turn will likely depend on the
stage of development the product is in. Don't feel limited by these
categories; they're provided only as rough guides.

Exploratory. These groups get at general attitudes on a given
topic, helping developers see how the eventual users of the product
will understand it, what words they use to talk about it, and what criteria
they will use to judge it. For example, a furniture company is
interested in what criteria people use to buy furniture and how they
buy similar items online. At the beginning of their development
process, they run focus groups and find out that, at first, people
insist on seeing furniture "in real life" before buying it (thus
negating their entire business plan for selling it online). Further
discussion reveals that it is only certain classes of products such as
couches and beds that are mistrusted without direct experience. With
other things (chairs, tables), most people have no problem buying based
solely on pictures and descriptions and, in fact, would prefer to do so.

Feature prioritization. These groups focus on what features
are most attractive to the group and why. They are held, in general,
near the beginning of the development cycle, when it's already clear
what the general outlines of the product are going to be. In these
types of groups, the assumption is that the participants are interested
in a certain kind of product, and the discussion centers on what kinds
of things they would like that product to do for them. For example, the
participants in a focus group for a homepage creation service were not
nearly as interested in community services as they were in tools to
help them build and market their own home-page. The "community feeling"
that the site was trying to communicate and the tone with which it
promoted itself meant little. For them the value in the site lay in the
tools and free disk space.

Competitive analysis. Just as it's important to know what
people value in the feature set that a product provides, it's important
to know what attracts and repels them with respect to competitor's
sites. Often held anonymously (with the commissioning client left
unmentioned), these focus groups attempt to understand what
associations people have with a competitor, what aspects of the
competitor's user experience they find valuable, and where it doesn't
satisfy their needs and desires. For example, a competitive focus group
of online technology news sites revealed that most of the content that
wasn't explicitly news was considered superfluous. Most people read
only one news site for a few minutes a day. What they valued most were
daily updates and links to other sites' content. Opinions, in-depth
background stories, and thematic collections of older stories were not
seen as valuable.

Trend explanation. When a trend in behavior is spotted,
whether it's driven by survey responses, customer service feedback, or
log analysis, it's often difficult to determine which of the many
potential causes are primary. Focus groups can help explain the
behavior by investigating the users' motivations and expectations.
These types of focus groups are generally held either as part of a
redesign cycle or in response to specific issues. For example, a survey
showed that when people used search services, they would use two or
three different ones with no particular pattern, but most people used
Yahoo! regardless of what or how many other services they used. Focus
groups with people who had taken the survey were called, and it was
discovered that Yahoo! was consulted first because people felt that it
did not contain much, but when it did, its results were of a
significantly higher quality than the competitors. Thus, it made sense
for them to always check it first, just in case.


ON: Focus Group via @jpenabickley