The Art and Science of Creating Effective Shared References: The Simple Secret to Increasing Conversion and Reducing Misunderstandings

By Eric L. Reiss

Imagine I am holding in my hand an ordinary 60 watt lightbulb. The bulb is made of glass and is roughly 10cm (approximately four inches) tall including the aluminium “E27” screw-in base. This stands for “Edison 27mm,” which was the standard introduced in 1909 by Thomas A. Edison and licensed under General Electric’s Mazda trademark. In other words, an ordinary 60 watt lightbulb.

Now close your eyes. Can you see my lightbulb in your mind’s eye? An ordinary 60 watt lightbulb. Got it?

Great. So, let me ask you some questions.

What color is it? Is it frosted or clear? Is it 110 volt or 220 volt? Is it broken or burned out?

Chances are you had a fairly clear picture of a lightbulb in your mind but didn’t ask these questions. That’s because there is no common conception of “ordinary.” What represents an “ordinary 60 watt lightbulb” to you might not be the same as that imagined by another reader. What’s more, I tricked you into thinking you got much more detail than you actually received—all the stuff about the E27 screw-in base is true, but it lulls you into thinking you have a better idea of what I’m holding than you actually have.

Quite simply, we lack a shared frame of reference.

Marketers and developers love to talk about obscure features yet often sacrifice important basic information along the way. One hears them cry, “Oh, we don’t have to say that. Everyone knows that already.” Sadly, the communication suffers as a result, so don’t let them talk you into this nonsense. In other words, don’t waste time (and space) explaining the history of the E27 base if this detracts from other important details.

Lack of shared references reduces conversions (people don’t buy things blindly), increases the need for help-desk assistance, and can dramatically decrease customer satisfaction. This article will (hopefully) get you to acknowledge the problem and start to review your own work in a new light. When you start to think about this, you’ll be amazed by the head-slapping moments you’ll begin to identify—from deciphering the cryptic messages you receive from public institutions to understanding cryptic menu options at a fancy restaurant.

The 80-20 Rule of Usability

Having observed usability tests for over 30 years, I’ve come to the conclusion that failing to establish a firm, shared frame of reference with users accounts for more than 80 percent of the problems that I see. People are confused. They don’t really understand what they are reading, there are questions left unanswered, or they simply don’t know what the writer and/or designer intended for them to do. Worst of all are sites, apps, manuals, etc., that simply fail to recognize the user’s true needs at any given moment.

The other 20 percent usually relates to technical glitches such as broken links, bits of code that are fighting each other in some hidden backend system, servers and other hardware that fail to work as designed, and so on. In many instances, one can argue that these problems, too, are related to a lack of shared reference—not between humans, but between machines. And as we enter the age of artificial intelligence, these problems will only escalate if we don’t start doing something about them now.

The Four Keys to Creating Great Shared References

Building solid shared frames of reference isn’t difficult, once you see the problem. This is what we, as writers, need to do:

  • Don’t take anything for granted. Not everyone understands things as well as you do. And even if they do, they always appreciate the confirmation of knowing that you understand things, too. Complacency is a big issue in technical writing today.
  • Answer the questions that you expect people to have. Make sure you don’t skip any of the basic information they will need. Strangely, many companies have a very poor idea of how their products and services are actually being used. If you don’t truly understand a user’s needs, you will never answer their questions satisfactorily.
  • Anticipate the questions they didn’t think to ask! You can check this by reading your text aloud to a friend (but not necessarily a subject expert). Have them ask questions. Probe. Ask them questions, too. You will often be surprised by what they come up with!
  • Keep in mind the specific communication environment. This will provide context for your message—context that either requires a more detailed explanation or context that permits you to make some reasonable assumptions.
“What” and “Why” Are Key Questions to Answer

In addition to the fairly obvious “What are you trying to tell me?” the second big what question asked by users is “What do you want me to do?” Although we writers and designers may have a clear intention, if users do not share this understanding, things are going to go horribly wrong.

I found this rather amazing text on Samsung’s website in roughly 2005:

We may send information on products and promotions in conjunction with our business partners. Please check this box if you do not want to receive this? Yes, keep me informed of the latest news on Samsung products, special offers, contests with fabulous prizes, and events.

This curious text was followed by just a single checkbox. What do you think will happen if you click it? (Personally, I’m impressed that they managed to work in so much confusing messaging and marketing blather in so few words.)

“Ah, but that was years ago,” I hear you cry. “We wouldn’t do this today.” Well, here’s an example circa 2019 from a major U.S. airline: “Do you want to print or receive your boarding pass by email?” I was given two choices: “Yes” and “No.”

Hey, this isn’t really a yes-or-no question—it is phrased as an either/or question. Clearly, I want a boarding pass, but is the site going to help me print it, or do I need to download it and print it from one of my own applications? I don’t know.

Without a shared understanding of “What,” things are never going to go well.

And that brings us to “Why.”

“An error has occurred. Please try again.”

We’ve all experienced messages like this when submitting information of some kind, but we don’t know why this happened. Did we make a mistake? Did the server make a mistake? A little extra information would help us understand. After all, if we made the mistake, there’s no point in repeating it, so give us some guidance. Again, we lack a shared frame of reference with whomever/whatever sent us this message. Surprisingly, these kinds of “microtext” are often left to developers instead of being reviewed and edited by professional writers.

There’s a wonderful sign in one of Copenhagen’s many public parks: “Please keep off the grass. Walking over this area will damage the flower bulbs.” This lovely message explains the “why”—it’s not just because some nasty official wants people to stick to the paths; it’s to preserve a wildflower area in a park that pretty much lets you walk anywhere you please.

Fear. Uncertainty. Doubt. FUD.

FUD is always a problem we want to avoid in technical communication—like the fear on the part of users that they will break something or lose data, the uncertainty that the action they are about to take will actually help them accomplish their goal, or the doubt that whatever they are about to do will truly solve their problem. Without a proper shared frame of reference, FUD flourishes and users are constantly frustrated and frequently angry.

Text, Images, Sounds—The Three Tools in Our Kit

Obviously, text is incredibly important. I’ll assume you know how to write, and since you’ve read this far, you’ve probably figured out the gist of my simple message. It’s all rather “in your face” once you recognize it.

The problem I often encounter is that someone higher up in the organization has decided that “People don’t read on the Web.” Actually, they do. Or, “Text should only be 10 lines.” Er, no. The King of Hearts from Alice in Wonderland knew exactly how long a text should be: “Begin at the beginning. Go on until you come to the end. Then stop.” It’s really no more complicated than that!

Images—photos, graphics, videos—can be a huge help in explaining things that are difficult to put into words. Although pictures are rarely worth 1,000 words, they can help. You can use photos to show how an unusual device is used (for example a vacuum cleaner designed to be worn like a backpack), or how large something is by placing it near an object of a generally known size (a hand, a common coin, a piece of fruit, etc). Remember, though, that pictures and text supplement each other; this is not an either/or situation.

I am amazed at how many companies forget the importance of these shared-reference images. For example, I live in Denmark and need a Danish keyboard on my laptop. But the photos on most websites are of American keyboards. Alas, when incorporating the three extra Danish letters, Æ, Ø, and Å, sometimes keyboards are rearranged in ways that are disastrous for touch typists. And that is why I am always forced to buy laptops I can actually see in person in a store. Ugh.

Sounds may or may not be important. Whether you need to highlight these is up to you, depending on your audience. But if a noise level or a notification sound is critical in some way, you’ll probably want to let your users/readers/customers hear things for themselves. As always, it’s about creating a solid shared reference.

If Content Is King, Context Is the Kingdom

Today, we are approaching the age of “Information 4.0.” Basically, that means that chunks of content are being chopped up into smaller chunks that can be recombined to provide a more “personalized experience.” Sadly, each time you cut a piece of content into smaller pieces, you invariably lose some of the context. Providing this additional information will be one of our great challenges in the coming years. And that brings us to metadata.

Garbage In, Garbage Out

For machines, “context” is provided through metadata—data about data—such as keywords and other tags that help describe a content chunk. Thus, the more you chop up something, the more metadata is needed to make sense of the pieces.

Let’s be honest, creating metadata is pretty boring, yet without it, so-called “artificial intelligence” has no chance of understanding how to deal with all of those pieces. The AI engine simply lacks the shared frame of reference needed to make sense of all the bits and pieces. As a result, we often see very odd information. My favorite is an airline seating plan I found where the seats are from one type of plane and the outline is for another type of plane. As a result, some of the seats are actually shown outside the fuselage!

Clearly, these are two pieces of information that don’t belong together. It would seem that the airline substituted a different plane for this flight segment, but the electronic mechanisms that displayed the seating information didn’t know to swap both content elements—they just changed the seating chart.

Because the creation of metadata is so time-consuming, eventually, artificial intelligence will start tagging things automatically. Great. But we, as humans, will have to maintain oversight. For example, I once found a sign for an English pub called “The Squirrel” with a big image of a rooster. I’m sure some pub owner thought this was very funny. However, if the machine thinks that a squirrel and a rooster are the same thing, we’re going to run into problems. Heading off these problems is the second challenge we will have to face in the era of Information 4.0.

In the end, it all boils down to building effective share references. It’s a remarkably simple concept when you think about it. And hopefully, you will, because your professional technical communication efforts can truly help make the world a little less frustrating.

ERIC REISS (er@fatdux.com) has held a wide range of eclectic jobs: piano player (in a house of ill-repute), senior copywriter (in an ad-house of ill-repute), player-piano repairman, adventure-game creator, and stage director. His experiences have served him admirably as a designer, content strategist, information architect, interaction designer, and usability “expert”—although he can’t explain exactly how.

In more mundane lives, Eric has been a two-term President of the Information Architecture Institute and Professor of Usability and Design at IE Business School in Madrid, Spain. Today, Eric is CEO of the FatDUX Group in Copenhagen, Denmark, a leading UX company with offices and associates in over a dozen cities worldwide. He also has several books to his credit, including the best-selling Usable Usability, which is now available in five languages. You’ll find him on Twitter at @elreiss.

Add Comment

Click here to post a comment

Download the Mar/Apr 2020 PDF

2020 PDF Downloads