• September 1, 2011
  • By Melanie Polkosky Human Factors Psychologist & Consultant - IBM/Center for Multimedia Arts (University of Memphis)
  • Interact

If It’s a Tornado, Just Text Me

Article Featured Image

Picture this: I’m cramped into a first-floor windowless bathroom with two screaming preschoolers on a mountain of pillows and blankets, and every phone we own starts to ring. Not once, but over and over, adding to the blare of sirens and the weatherman’s announcement that tornadoes are touching down a couple of miles away.

Through this spring’s historic storm activity in the U.S., I became an expert at receiving automated messages. As many as 60 texts and automated calls received almost daily for more than a month warned me of tornadoes, severe thunderstorms, floods, and flash floods. Even my 2-year-old, on hearing the ding of an incoming text, would holler, “Mommy, there’s a flood warning for Memphis.” 

Weather notification seemed like a good idea when I signed up, but it became an annoyance. It reminded me of a friend’s winter notification experience: All her phones would ring at 4 a.m. with the message that school had been canceled or delayed. Yet, a common refrain I hear from clients is, “Could we just leave a placeholder for messages at the top of the call?”

A recent USA Today online editorial, “Our View: Alert Wireless Customers on Limits,” said messaging decisions may hold a business benefit. Sure, automated messaging can be useful, but it must be relevant to users. And this is where most automated messages violate a basic communication rule.

Relevance theory suggests the act of sending a message implies the message is worth listening to. Yet most aren’t. Businesses assume the mere presence of their message transfers information and triggers a desired recipient action. But ubiquitous poor messaging has conditioned people to believe that most automatic messages lack specific relevance to any one user, especially me. As an example, my pediatrician’s office has had the following message on its IVR for seven years: Please listen carefully as our options have changed.

One of my oft-repeated recommendations to clients involves removing or redesigning extraneous and excessive messaging. Most messages apply to only a small subset of the user population and are often extremely dense, wordy, or jargon-filled. Automated messaging, when designed well, puts a significant onus on the sender to make the message relevant. The sender should know highly specific information about each unique recipient, correctly anticipate his information needs, and send a targeted communication. This requirement should conjure images of whirring databases and scores of interconnected technology to support such precise communication. In reality, that’s exactly what’s lacking. Instead, we typically broadcast a widely applicable, generalized message that’s meaningless to most individuals.  

Defining relevance is a context-dependent, yet critical, issue to address when designing outbound notifications or any other type of automated messaging. Important design considerations include:

1. Means of delivery: Does the user want the message delivered via email, phone call, or text? Is she likely to be doing something else when the message is delivered? What is the most accessible form? Does the means of delivery change with the content?

2. Urgency: Does the message concern physical well-being or a threat of harm? Are multiple notices warranted or appropriate? Does the user expect the message? Is immediate action required?

3. Content efficiency and response: What is the expected user response? What is the critical information for users? Is this information actionable? Is related content necessary? Does secondary content obscure critical data? Is there a more efficient delivery method (e.g., audio icon)?

4. Timing: When should the message be delivered? Is the information tied to another event? How far in advance should it be delivered? Is a single message sufficient? When should messaging stop?

5. Specificity: To how many users does this message apply? What technical capability do we need to make this information specific to groups of individuals?

Two examples of weather messages illustrate the difference between texts that do and don’t take these design considerations into account: 

• Current: A severe thunderstorm warning has been issued for Memphis.

• Relevant: A severe thunderstorm with hail and tornadoes will hit your location approximately 8 minutes from now. Please take cover immediately.

Relevant message design requires a highly user-centric perspective, like many other types of automated communication. The next time you’re considering creating a message you think is simple, ask yourself: How useful would this specific message be to me in the context of my own daily life? If it simply adds to the cacophony, rethink the message.

Melanie Polkosky, Ph.D., is a social-cognitive psychologist and speech language pathologist who has researched and designed speech, graphics, and multimedia user experiences for almost 15 years. She is currently a human factors psychologist and senior consultant at IBM. She can be reached at polkosky@comcast.net.


SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues