UI Design Newsletter – August, 2009

Unintended consequences from standards?

Every once in a while, I read a journal article that represents a heck of a lot of research effort, and wonder, how did they reach that conclusion from the evidence? For example, Kristin Eschenfelder explored how web standardization projects contribute to conflict in Web design.

After 20 years experience helping organizations create effective application and web standards, I read this with some puzzlement. After all, doesn't a design standard settle conflicts by saying "this is the best answer to the conflicting demands on navigation, layout, vocabulary, and branding"?

Doing forensics on three sources of conflict

Each company was substantial, with 5,000 employees or more. She asked the Web managers to "describe problems they experienced in coordinating Web management work with others in their organization". (To avoid "leading the witness" she did not mention "standards" or "information architecture" or "design".)

Her findings emerged from a systematic content analysis of these interactions. She found three sources of organizational conflict:

1. Battles about the information architecture for home page links as well as second and third level menu categorization.
2. Conflicts on Web page templates including color, layout, logos, graphics and global navigation.
3. Arguments surrounding the content management system (CMS) database entities, attributes, and position of fields on templates (only one of the four companies used a CMS).

At this point, pretend you are watching the usability version of the television hit "CSI" (Crime Scene Investigation). Pretend you are a "usability forensics expert".

What caused these organizations to have conflicts from "standards"?

Digging into the findings

Interestingly, Eschenfelder's content analysis of the conversations yielded additional insights into why the standards seemed to have fallen short. Note the word "seemed".

In summary, many factors conspired to put the blame on "standards" while the actual blame should fall on lack of a strong usability methodology. If your appendicitis operation fails, do you blame Modern Medical Practice, or do you blame the doctor and his operating procedures? Probably the latter. Let's take a look.

Who are the customers?

Eschenfelder found her 18 participants discussed "ambiguity about customer needs". This was manifest in issues you and your team have probably met as well:

1. A single organization served different customer groups. The perception of customer needs and expectations varied across sub-units in the organization.
2. Upper management had failed to prioritize which user groups were most important.
3. Lack of direct interaction with the customers. (Sound familiar?)
4. Organizational uncertainty regarding customer willingness to purchase online vs. traditional brick-and-mortar channels.

Consequently, as Eschenfelder says, "If it is not clear who the customers are, and what needs they have, it is difficult to develop consensus on how to design a classification scheme to help them."

Sound familiar?

Here's a big clue for your usability forensics. Eschenfelder quotes how a home page manager depicts the bickering among different business groups:

"As an individual business unit becomes more aggressive and understands [that] one of the ways that people are coming into their site is through [the home page], they're going to want to control that, too... So that's what I get challenged with. How come one business is listed ahead of another business?"

Usability methodology solves the problem

Let's cut to the chase. You probably felt from the quote just seen, that design "standards" were more of a battle of wills or persuasive eloquence by some manager.

But what about a "usability approach"? We know the answer. If the interests of the organization as a whole were considered, the standard should incorporate knowledge of the user types as well as the relative profitability of the various lines of business.

We didn't see any of this systematic thinking when hearing Eschenfelder's depiction of the problems with standards.

So now you know. The problem was not the standard. Instead, it was a lack of usability methodology that normally would support the correct design of a standard.

Your forensics solution

Let's take a look at how usability methods solve the problem of conflict.

1. How do you know who your customers are?
You conduct research on user profiles. Get your marketing data on who uses your product or service. From there you conduct interviews to identify clusters of people who share similar goals, skills, and task requirements. We call these clusters "personas". A single "persona" captures the flavor of a user type. You might end up with 3 to 5 personas that allow you to rapidly evaluate how your design gets interpreted and utilized. So you need personas to help make sure your standards work.

2. How does your organization make money?
Some personas may be more profitable than others. You should prioritize which persona gets the biggest chunk of attention. Or, alternatively, you could "segment" your website to provide functions for different groups among your personas.

One example comes to mind regarding selling auto insurance. Perhaps your research shows that some consumers like to shop on the web because they like to do the research. They also feel qualified to compare plans. Meanwhile, a different "segment" of personas might prefer to delegate these tasks to an agent.

Voila, you steer the researcher personas toward a website comparison tool, and the agent personas towards their nearest agency. (Check out the Web regarding landing pages for "audience segmentation".)

Both of your "forensic solutions" given here contribute to making a standard that resolves conflicts in your organization.

The standard pay-off

As Eschenfelder concludes her study, she reminds us what standards contribute to any organization:

1. Reduced support costs
2. Easier development and maintenance of websites
3. Customers get a uniform set of cues that support navigation and scanning.

First lesson on usability forensics

And as you enter your new career as "usability forensics expert" you have learned a few things too.

1. Don't expect a standard to work unless you really understand "usability" as a methodology.
Don't practice surgery until you complete your MD and internship.

2. "Institutionalize" usability.
Have your organization commit to a set of personas and the research required to get them right. Make sure colleagues understand your discipline.

3. Understand the return on investment (ROI) from each persona.
Decide how to handle each persona whether it's using the Web, phone, brick-and-mortar, or all three.

All in all, prepare to segment your audience. Design your standard to include landing pages that let visitors select their interest. Then, based on their selection, deliver the types of pages best suited for their vocabulary and intention.

Watch television

Usability offers a lot. But your intelligence offers more. Watch CSI (Crime Scene Investigation) and consider how being a scientific sleuth applies to your job.

Maybe sometime soon we really will have a job title "usability forensics". Today's study was your first lesson.

References

Eschenfelder, K. The customer is always right, but whose customer is more important? Information Technology & People, 16(4), 419-439, 2003.