Mind the Gap -- On the appropriate use of focus groups and usability testing in planning and evaluating Interfaces

By Kath Straub, Ph.D., CUA, Chief Scientist of HFI.

Focus groups and American juries

Have you ever been on an American jury? The objective of a jury is to examine the evidence in a given trial and determine whether the evidence is compelling. Within the deliberation process, members of the jury must come to some level of consensus.
Often consensus is achieved by successive approximation:
discussion, opinions get closer, discussion, opinions converge even more, until eventually the required consensus is reached.
Juries provide fascinating environments to explore the dynamics of human interactions. Within a jury, there is usually a natural leader. The natural leader is usually someone who is interested in justice and/or the law. They may also feel responsible to discharge their civic responsibility. Critically, however, the natural leaders also typically come with an outgoing personality.
They are talkative. Articulate. Maybe even impassioned. They are often able to sway others on mere conviction.

Focus groups are a lot like juries.

The focus group: What is it good for?

As a rule, usability practitioners don't do a lot of focus groups.
Clients are often surprised to learn this.

Focus groups do have a place in the design lifecycle. It's just that the time for focus groups is different from the time for usability testing. Usability testing is for evaluating in the here and now. Focus groups are for innovating and looking forward.

Focus groups have a place very, very early in the product innovation process. They are appropriate for exploring unknown territory. This is territory where stakeholders do not yet have a clear idea of what their target customers want or need. It may be territory where stakeholders seek to fill a known (but largely
misunderstood) gap in a product line. To this end, focus groups provide a good strategy for defining new products, systems or functionality or, subsequently, exploring user attitudes about a proposed new product or tool.

Focus groups are also helpful when an industry is undergoing a paradigm shift. In this case, they offer one in a set of research strategies to help designers quickly gather data about how target users interact with, use, and understand certain types of information or products. This understanding, then, provides the basis for visioneering... brainstorming...innovating new directions. Sometimes this innovation happens within the focus group. Other times it happens as a result of the focus group.

Essentially then, focus groups are excellent for quickly eliciting new directions, new ideas, enhanced functions, or different directions.

Usability practitioners don't do focus groups because typically they are no brought in early enough in the development cycle. Instead, the input of the usability group is solicited only after the commitments to functionality are made and the development is in progress. At that point the usefulness of focus groups has faded. Alternately, we are asked to evaluate and enhance existing interfaces. This requires evaluating and improving. Not innovating.

And we've got stories...

Companies that have institutionalized usability actually bring usability practitioners into the process early enough so that using focus groups is appropriate. As more companies reach this level of usability maturity, more practitioners should be called upon to do this kind of work. Usability practitioners conduct focus groups when they are given the freedom to truly evolve a design. That is, to extend it, re-position it, and re-align it at the foundational functionality level. To identify new opportunities to delight the user in new ways. THEN, if the situation calls for it, strike up a focus group.

Most times, however, even if a site is slated for a complete overhaul, the functionality is not going to change fundamentally. In that case, the return on investment is higher if other user- centered research strategies are exploited.

There are many organizations that say they use focus groups as a means to conduct usability testing. I'm never quite sure what that means. Focus groups are not an effective method for this task. There are three big reasons that focus groups are not useful for evaluating the usability of an existing site.

Focus groups and juries: Redux

(1) The social dynamic of even a well moderated focus group session can give rise to inaccurate data.

Charismatic participants able to articulately state their position (independent of its logic or validity) can dramatically sway the beliefs of the entire group -- at least temporarily. This dynamic can lead other members of the focus group to provide false assent or feedback. Feedback moderated by a charismatic participant rather than a participant's own beliefs and thoughtful conclusions is not helpful. (Note that this problematic dynamic also holds for American jury trials). Further, the context of a focus group -- everyone giving opinions -- can make participants feel pressured to have an opinion even if they don't know or really don't care.
Explaining to participants that not having an answer actually is an answer is always tricky.

Do I do what I say or say what I do?

(2) What people say they want/will do is often different from what they need/do.

Both anecdotal evidence and careful scientific study demonstrates that actual behavior often differs from predicted behavior or expressed desires.The classic example of this problem is the Ford Edsel. The Edsel design was derived from a long series of focus groups -- more than for any other automotive product.
The focus groups addressed topics ranging from color to tail size to reasonable cost. In the case of the Edsel, it is clear that what people thought they wanted was very different from what they selected and paid for in the show room.

Mind the gap

(3) Most importantly, traditional focus groups do not provide an opportunity to directly observe users completing tasks.

Focus groups provide no means to directly explore any match (or mismatch)between the site model and the user's mental model. Think of it this way -- if your goal is to improve how well your application or Web site works, you need to do two things: figure out what doesn't work and make it work. Focus groups are about talking about what works and what doesn't, and what might and what might not. Usability testing is about collecting the user- centered data to figure out how to fix it.

Time for testing

Usability testing provides a type of data that sophisticated practitioners can leverage to evolve the application/site to better reflect the user's conceptual model. When users struggle on a site or application, this indicates that there is a mismatch between the user's mental model/expectations and the way that the site works. Done well, usability testing can provide key insights into the user's conceptual model for the tasks explored in the testing session. This data allows the practitioner to examine points of mismatch and to re-engineer the task flow/information presentation/visual hierarchy... so that it is consistent with the user's expectations.

So the goal of usability testing is to collect data that is specific enough not only to identify challenges that users face, but also to describe expectations or their mental models. Collecting data at this level of granularity requires directly observing representative users complete critical tasks on the site/application in a 1-on-1 interview format. Throughout a usability testing project, specialists should work to systematically identify and prioritize points in the task flow or information search where users slow or stray from the optimal (or intended) path. This approach allows the tester to systematically identify mismatches between the site's navigation structure, information architecture, and detailed design and the users conceptual task model, information hierarchy, and expectations for use. Usability testing also provides a means to evaluate the appropriateness and usability of functionality and content. Finally, usability interviews can also be used to collect data to inform decisions about the effectiveness of look-and-feel/ branding and the appropriateness of graphics on a site. (E.g., If you have a site about a very dire illness you probably don't want shiny, happy people dotting the page.)

Enough small talk

The method you need to use depends on the goals of the exercise.
If you want to explore possible directions for increased functionality and test market interest, use a focus group. If you want to evaluate and improve the effectiveness of a current application or site, use usability testing..