Some time ago, I came across Boxes and Arrows, a journal concerned with design: graphic design, interaction design, information architecture and so on.

In the September 2008 issue, Jens Jacobsen talks about information architecture for audio.  As he points out, audio is linear. “You can only consume it in a linear fashion and you have to listen to it at a given speed.”

When beginning an audio-related project, ask yourself whether audio is the right medium for your message. In some cases, text is a better choice and in other cases it’s video. Don’t use audio just because you can. If you are certain audio is the best choice, there are several fields to help inform how you implement it.

Jacobsen offers several guidelines from different fields. It’s worth reading the entire post; these are simply highlights from different field — like information architecture:

• State the length. Let people know how long the audio’s going to be.
• Introduce the topic. “In printed text…[this] might seem hackneyed, but with audio… it’s best not to jump right into the topic.”
• Provide orientation from time to time. Let the listener know where he as and what’s coming next. In a long piece, consider giving an option to skip sections via the interface.

From journalism:

• Keep it short. (My opinion: because it’s one way, five minutes of audio feels a lot longer than five minutes of conversation.)
• Repeat often. Jakobsen means a summary at the end, but also repeat the main subjects or themes.  Don’t refer to them by pronouns or synonyms.  You know what you’re talking about, but your listener has nothing to go on but short-term memory.
• Take advantage of the possibilities. Change the style of speech, the tone, the speed.

I especially like the suggestion, “Don’t overuse the thesaurus.” If you’re calling people learners, don’t change them into users, then stakeholders, then students, then knowledge partners.
Suggestions from usability engineering:

• Design for the target audience. “Convince your design team to produce content for the users, not its creators.”
• Create personas. Represent your target audience in the audio.
• Create scenarios. You’ve got those personas sitting around waiting for osmething to do.
• Test with users.

On that last point: in my experience, there isn’t a lot of testing of audio for things like online learning. It’s as if having a professional voice (or [shudder] your boss’s boss’s voice) will overcome any shortcomings in the text. That’s the audio equivalent of believing that if you choose the right font, your text will be more understandable — as opposed to easier to read.

Although interface guru Jakob Nielson recommends usability testing with “only five users,” you can see from the chart in his article that it’s possible to benefit from tryouts with just two or three people.

There’s all kinds of audio in learning: voiceovers, audio as part of video, and plain vanilla podcasts. If you’re going through the trouble to deliver information via audio, it makes sense to think about ways to make the optimum delivery.  Boxes and Arrows is all about delivery.

1931: noted on-air personality Pope Pius XI, with (at right) founder of the Vatican radio system, Guglielmo Marconi.