How to Get Away with Murdering a Bad Idea

Finding Our Way

Navigating the practice of Information Architecture

A column by Nathaniel Davis
November 3, 2014

Business leaders, clients, marketers, and product owners typically underestimate the strategy, design, and development effort that goes into creating effective computing user interfaces.

Companies whose business model is Web based—for example, those that develop a Web product or service—may have a greater appreciation for information architecture (IA) and user experience (UX) strategy and design best practices. However, this is much less likely for other types of companies. Getting businesses other than technology companies to carefully consider Web user interfaces as a strategic extension of their business will require persistent efforts in education, building trust, and validation.

Champion Advertisement
Continue Reading…

Until this shift happens, many businesses will continue to oversimplify the delicate connections between a business offering, its customers, and the computing user interfaces that support sales and service. So be prepared to encounter clients who sometimes feel compelled to dictate constraints, as well as the approach to setting IA and UX strategy— thus, potentially, planting the seeds of bad ideas.

Whenever possible, kill bad ideas, nipping them in the bud—of course, with all due respect and professional courtesy. Fortunately, killing bad ideas is not a crime. In fact, when you can pull this off, you’ll be returning needed sanity into the universe—and that’s a good thing! So, in this month’s column, as Figure 1 illustrates, I’m going to show you how to get away with murdering a bad idea in three steps. First, let’s assess the scene of the murder.

Figure 1—The scene of the murder
The scene of the murder

The Crime Scene

It’s 9 am, and you’ve received a text message from your client. She wants you to join a meeting to discuss the latest plan. She’s communicated that it’s a “high profile” plan, and they’ve scheduled this meeting with a sense of “code-red” urgency. It’s happening, right now! So you decide to sacrifice your morning ritual of coffee and purging junk email and immediately proceed down the hall—unless you’re a consultant, in which case, you hop into your car.

When you arrive at the meeting, you notice that everyone has a look of resolve about them. In fact, you might characterize the vibe in the room—which is full of marketers, analysts, and business stakeholders—as oddly refreshing.

As you take your seat, you scan the cheery faces in the room, but suddenly pause when you see the face of your client who texted you. Observing her lifeless gaze, the pleasant feeling that you had quickly flushes out of your mind and body—especially when her eyes slowly roll up into her head and she nods with subtle disapproval that only you sense. As you realize that your client didn’t invite you to take part in some awesome project, the stench of a bad idea begins to tickle the hairs of your nose.

You discretely reach under your chair and adjust it—slowly locking it into the most upright position possible. You then trade in your cool, relaxed persona and, instead, adopt the posture of a sniper whose target is UX-design mediocrity, with one eye on the HIPPO in the room—who is the likely source of the highest-paid person’s opinion that you’re about to hear—and your other eye on everyone else. No word goes unheard.

Recognizing a Bad Idea When You Hear It

Whether you’re an information architect, UX strategist, UX design lead, or design ninja, listening is one of your most valuable skills. The more experienced you become, the easier this gets. You’ll discover that bad ideas have a conversational signature, as shown in Figure 2. Let’s return to the scene of that impending murder.

During the meeting, you hear all about what the business wants to do, how they think the user interface should behave, and when the business expects the code to go live. There’s no mention about the impact on users. There’s no concern about doing user-focused research to help validate their painpoints or expose product opportunities—and their measures for success are shallow, at best. Yet, somehow, everyone in the room seems to agree about how to make these improvements.

Figure 2—The signatures of a bad idea
The signatures of a bad idea

How to Dispatch a Bad Idea

If you have anything to do with it—and you will—the idea that they just proposed should be dead on arrival. Improving your chances of getting away with its murder will require you to be like Annalise Keating, the lead character in the new ABC series, How to Get Away with Murder.

Annalise, played by Viola Davis, is a law professor and practicing lawyer who takes on the most questionable clients—often hardened criminals. Her formula for getting her clients off the hook resembles what IA and UX professionals must do to promote the adoption of sound design strategies. In the first episode, Annalise described her three-step process with pure, unapologetic certainty:

  1. “Discredit the witnesses.”
  2. “Introduce a new suspect.”
  3. “Bury the evidence.”

She finishes her list with this great line: “Throw so much information at the jury that they walk into the deliberation room with one overwhelming feeling of doubt!” Now, let’s review how to get away with sewing seeds of doubt about a bad idea.

Step 1: Discredit the Witnesses

Translation—Broaden the range of possible assumptions.

While you should never recklessly aim to discredit your client, do challenge their assertion that a digital strategy is purely quantitative or circumstantial. Instead, extend your client’s data-driven motivations by introducing qualitative factors—filling the obvious gap that the first item in Table 1 identifies.

Encourage your client to consider the possibility that more specific causes for an issue exist—such as the performance of a questionable interaction with information or user interface (UI) behavior. For example, it might not be the entire page that is at issue. Sometimes it can be a single element like the visual affordance of a button or maybe the use of language in the body text that doesn’t align with user expectations. Or it could be a much larger issue such as the organizational structure for the site’s content and the labels of items in the navigation. These are just a few examples of cases where you can validate the speculative results of data analytics through contextual probing that yields more conclusive data.

If there’s no time to consider more granular and qualitative aspects of the user interface, present this as a risk to any future analysis. Next, recommend a future effort to establish key performance indicators (KPIs) that would help in measuring the integrity of the user interface. Then, in the future, you’ll be able to use UI-based metrics to make IA- and UX-based decisions for this data-driven client—even under tight time constraints.

Key recommendation—Encourage your team to probe the qualitative performance of both UI elements and a Web site’s structure.

Step 2: Introduce a New Suspect

Translation—Show how there are two sides to every story.

Unless you’re working with experienced product leads, it’s most likely that your clients can see only with their own eyes. As a consequence, their suspicions—about how to approach a digital strategy—will likely be business centered, marketing centered, or operations centered. While this is only natural, it’s also the fuel that can easily ignite bad ideas—and it shows apathy regarding IA and UX strategy.

Not considering impacts on the people who engage with and on behalf of the business can compromise a sustainable digital strategy. To avoid this, help your client to realize:

  • Their strategy must correlate to specific user scenarios.
  • The value of any effective user interface is reciprocal—that is, all sides benefit.

Start by encouraging your team to map each business goal to a thoughtful model of user behavior. This model should help the team to analyze the personal dispositions of a site’s or application’s users, who will be affected by their business strategy. The persona is the industry’s most common modeling approach. Personas include, but are not limited to

  • evidence-based and deduced assumptions about the target audience
  • the relationships that users expect to have with the business through the use of information technology
  • users’ informational needs
  • demographic data about users—such as age and influential cultural factors and habits

In every case, an objective analysis of the people who will actually use the computing user interface that you’re creating will paint a very different picture. This is the type of disruptive provocation that can eventually steer your clients away from bad ideas and toward the formation of a strategy that is based on a sound foundation.

Key recommendation—Impress upon your clients the need to correlate business KPIs with related customer-focused KPIs.

Step 3: Bury the Evidence

Translation—Improve your client’s assumptions by providing better sources of information.

Bad ideas generally arise from a lack of sufficient information. Hence, knowing what areas of a user interface to diagnose for actionable insights is central to illuminating the gaps in a UX strategy.

The last nail in the coffin of a bad idea will be your ability to communicate the value of several proven methods that can help you to triangulate deficiencies in user interfaces. The following IA and UX research methods are essential parts of the toolkit of any UX professional:

  • heuristic evaluation—Compares UX strategy or existing user-interface functionality to current industry standards and best practices.
  • search-log analysis—Helps you to discover semantic nuances in language, gaps in content, and errors in the prioritization of content, and reveals unanticipated search behaviors.
  • user interviews—Helps to remove guesswork by acquiring direct feedback from actual users.
  • contextual interviews—Goes beyond what users say and looks at what they actually do in real-world scenarios.
  • secondary research—Provides an affordable and efficient approach to gaining valid insights by leveraging research from other credible sources.
  • usability testing—Offers the best way to test, learn, and improve a user interface before its release.

If you’re fortunate, you’ll have many of these information sources at your disposal when devising an information architecture, UX strategy, or design solution. However, if you’re like a large percentage of UX professionals who have to work within more modest budgets, simply get the most out of whatever information you have. The point is to use whatever research approach will help you to either validate or disprove any of your client’s assertions about the ways in which a user interface might impact the user experience and, thus, business objectives.

Key recommendation—Encourage your team to acquire the appropriate depth of evidence that is ultimately necessary to determining an effective IA or UX strategy.


Computing user interfaces serve as human touchpoints in interdependent ecosystems comprising businesses, customers, employees, and vendors. In many cases, the user interface actually is the market-valued product or service. In others, it contributes to the overall customer experience. Or it may do both. In each of these cases, the effort that goes into the overall architecture and design of a user interface would be the same. When developing products or services, you must consider the engineering complexity of user interfaces, the human factors of its users, and the use of research methods that provide insights that contribute to user-interface improvements.

Finally, by inserting the analysis of information architecture and user experience into your client’s overall business model, it’s possible to prevent most bad ideas that come from our business, marketing, and product development counterparts from being implemented. However, in this age of digital information, most product and service businesses are still playing catch-up, so need to understand what many information architecture and UX professionals already do. As a result, we have an opportunity to help both public and private organizations to integrate user-centered design approaches into the development of strategies for products and services—regardless of the digital or analog nature of the customer experience.

So, when you come face to face with the classic signatures of a bad idea, nip it in the bud by following the three steps that I’ve outlined in this column. Rely on proven IA and UX wisdom when setting strategy. Good luck! 

Founder at Methodbrain

Franklin Park, New Jersey, USA

Nathaniel DavisNathaniel has over 25 years of experience in human-computer interaction and is a leading advocate for the advancement of information architecture as an area of research and practice. He began practicing information architecture in the late ’90s, then focused on information architecture as his primary area of interest in 2006. He has made a study of information-architecture theory and how that theory translates into science, workable software, and methods that improve human interaction in complex information environments. Nathaniel was formerly director of information architecture at Prudential. His information-architecture consultancy Methodbrain specializes in UI structural engineering.  Read More

Other Columns by Nathaniel Davis

Other Articles on Design Process

New on UXmatters