Is 93% of communication made up of body language?

Ever heard the statistic that 93 percent of communication is made up of body language? It’s widely repeated . . . and it’s totally inaccurate.

The magical 93 percent number comes from a 1971 study by psychologist Albert Mehrabian. But this study was actually very limited. It asked listeners to judge a speaker’s mood based on the way he delivered just one word. Mehrabian himself has pointed out that, beyond the confines of his initial study, this statistic is totally misapplied.

Obviously, body language and other non-verbal cues make up a large part of the way we talk in face-to-face settings. In fact, we can communicate effectively using just our bodies, as anyone who has ever signaled for a drink across a crowded bar or blown a goodbye kiss through a window already knows.

Ever heard the saying that actions speak louder than words? It’s tempting to believe that, even when someone speaks untruthfully, his body will always inadvertently give him away. Some body language “experts” even attach fixed meanings to certain gestures. If someone folds his arms while listening to you, he’s angry. If someone lays his hand on your arm while he talks to you, he’s flirting.

In reality, gestures can be just as ambiguous as words. Some are even explicitly designed to be ambiguous. That “flirtatious” hand on the arm? If you don’t reciprocate by showing interest, it can easily be dismissed as friendly or even accidental contact.

The truth is, if body language really were so persuasive and interpretable, it would be far more useful. We would be able to communicate across language barriers using only non-verbal cues. We’d know instantly when politicians were lying, for instance, just by studying their non-verbal cues. But we simply can’t do that accurately.

So where does body language fit into conversation analysis? In reality, we communicate in a multi-modal manner, both verbally and non-verbally. We can tell a friend about a date gone awry and hail a taxi at the same time, through a combination of speech and action. It’s best to appraise body language in this same multi-modal context. Non-verbal cues, like folded arms or downcast eyes, make up part of the conversational picture, not the whole picture.

It seems actions don’t speak louder than words. But the words we choose can affect the actions and responses we produce.  Hotels often encourage guests to reuse their towels with a sign in the bathroom. But signs that appeal to guests to reuse towels for environmental reasons have proven ineffective. A better sign appeals to social norms, saying something like “Most guests choose to reuse their towels.” Changing the wording of the sign is all it takes to produce a change in guests’ actions.

This is a classic example of choice architecture, the process of designing the ways in which choices are presented to others in order to secure your desired outcome. Choice architecture can be easily incorporated into your conversational tool kit.

If your questions are being met with “no” when you want them to be met with “yes,” you might just need to change the choice architecture at work in your request. Even changing one tiny word can create big results, as a 2007 US study demonstrates.

The study was built to address an endemic problem in doctor-patient communication. Patients often come to the doctor with more than one concern, but can be reluctant to raise multiple issues within the framework of a single appointment. As a result, they can leave their appointments feeling dissatisfied – or, worse, with untreated health problems.

Even when doctors asked, “Is there anything else I can help you with today?,” the study found that only 50 percent of patients brought up another issue. Did the problem lie in the patients’ reluctance to discuss their health – or in the way the question was designed?

Well, when doctors asked, “Is there some other issue you’d like to address?,” a staggering 90 percent of patients responded with yes.

Both questions were closed, meaning they invite a yes or no answer. Why did so many more patients respond positively to the second question? The answer lies in the use of “any” and “some.” The word “any” is negatively polarized, meaning we use it more frequently to describe a negative (“I don’t have any friends” as opposed to “I have some friends”). Questions with “any” are easier to answer with “no.” That’s why queries like “Any other business?” or “Any questions?” are so often met with awkward silence.

Questions that use “any” invite a negative response and shut down further interaction; questions that use “some” have the opposite effect. Switching “any” for “some” might be all you need to do to change a negative answer into a positive one.

Ask and you shall receive . . . but what you receive might depend on how you ask for it.

Every day we use talk to get and give help, to receive and provide services. And it’s talk that can provide the key to getting and giving better service.

As a general rule, we don’t ask for what we want outright. Requesting help can make us feel vulnerable; asking for what we want can feel too direct. In fact, in everyday conversation, most people avoid asking direct questions.

Sometimes we can engineer the service we want without asking for it directly. Often, instead of a question, we make a statement, like “I’m hungry,” that is designed to prompt an offer, like “I think there’s some leftover pizza in the fridge.”

When we do ask questions, the way in which we design them can reveal how entitled we feel to receive the service we’re asking for.

A request along the lines of “Could I make an appointment for tomorrow, please?” is clear and direct, using a simple modal verb to communicate the question. It creates the impression that the speaker feels entitled to the service.

By comparison, a request like “I was wondering if it would be possible to make an appointment for tomorrow, please?” is indirect. The request is communicated through a series of apologetic phrases. This creates the impression that the speaker is unsure whether she is entitled to the service or not.

What’s the lesson here? The clearer the question, the stronger the sense of entitlement you create – and the more likely you are to receive the service you’re asking for.

Understanding the way we ask – or don’t ask – for the things we want also provides insight into how to give better service.

Making direct requests can make people uncomfortable. The fewer questions a client needs to ask in a service context, the better! So, giving great service means anticipating questions and addressing them. In particular, it means looking beyond the form of the question to its function.

When a customer in a hotel bar asks whether Wi-Fi is available, she’s actually concealing a number of further requests within her initial question: Do you have Wi-Fi? How can I access it? What is the password? Effective service anticipates the requests concealed in her original closed question and meets them.

Have you ever made a half-hearted offer that, to your chagrin, your conversational partner accepted? Have you ever felt forced into making an offer you really didn’t want to make? Making and accepting offers can be a conversational minefield. Here’s how to negotiate it.

Accepting offers can require delicate conversational strategy. If you’re accepting an offer, how can you do it graciously? Try these two simple strategies.

Let’s say someone offers to buy you a coffee. It’s fine to accept, as long as you don’t do so too speedily. Pausing before you accept will show that you’re not taking the offer for granted.

A more involved offer, like the offer of a lift home, is best accepted through a sequence of brief conversational turns. If someone offers to drive you home, rather than immediately accepting, use your next turn to check that the offer is genuine and not made from mere politeness. Do this by saying something like “Are you sure it’s not out of your way?” With this follow-up question, you create the chance for your conversational partner to withdraw his offer.

Making offers, too, can be complicated. It’s one thing to make a genuine offer, but we can often feel pressured into offering to do things we don’t want to do. When our offer is accepted, we feel frustrated and resentful. Why do we make these unwilling offers in the first place? It might come down to the tactics of our recruiter.

A recruiter is someone who, through conversation, places others in a position to make an offer. There’s nothing wrong with that in principle. However, serial recruiters often force offers from others by placing a mental burden on them. Here’s a typical request from a recruiter:

“Let’s make a time to chat about Alison’s surprise party.”

This conversational offer sets up an expectation – in this case, that a meeting needs to take place – but places the burden on the recruited to drive the suggestion forward so the objective of the conversation can be met. If you’re dealing with a serial recruiter, gently turn the burden back on the person by nudging him out of this conversational pattern. Instead of responding with an offer – like “Okay! Why don’t we meet at my house next Saturday?” – respond to the request without driving it farther forward by saying something like “Sure! Let me know what times work for you.”

Once you’ve identified them, the problems around making and eliciting offers are easy to avoid.

If you’ve ever worked in a customer-facing role, you’ve almost certainly endured HR-mandated conversational role-play designed to improve corporate communications strategy. But sitting across from your boss playing the role of “irate customer” isn’t just awkward. According to conversational analysts, it’s actually counterproductive.

Role-plays and simulations are classic communications tools, but they don’t work. Why not? Because they create an artificial conversational context.

Remember that a conversation is a collaboration in which two or more people work to complete a project through a series of conversational turns.

With this in mind, let’s compare the project in a role-play context and a real customer service context. In the real context, the employee’s project is to deliver service smoothly and efficiently. In the role-play, the employee’s project is very different. It might be to complete a training module, secure a promotion, or impress a manager. In a real context, the employee draws on her own conversational tool kit. In a role-play, she tries to match her conversational turns to policy, or even a customer service script.

In fact, when conversational analysts compared conversational role-plays with actual conversations, they found several critical differences in content and tone. Here are just two findings made by conversational analysts when comparing role-plays to real contexts.

In the context of police training, for example, role-playing the interrogation of suspects is a common exercise. Often, interrogators will ask open-ended questions, like “What were you doing on the day of your arrest?” That’s because written guidance instructs officers to avoid asking closed questions in an interrogation. In fact, conversational analysis of actual interrogations found open questions could sometimes elicit answers that were too broad or off-topic to be useful. Closed questions, meanwhile, like “Could you tell me what happened before your arrest?” were often more effective at eliciting useful information.

Meanwhile, neonatal doctors are instructed to make recommendations to parents of premature babies by using the phrasing that they are “in the best interests of the child.” While intended to sound reassuring, in reality, analysts found that this phrase shut down opportunities for questions – which, in turn, led to conflict between parents and medical staff.

Leaving the phrase out opened up the possibility for parents and medical staff to discuss the recommendations further. Parents felt their concerns were being heard, and conflict was avoided.

Role-plays create artificial insights into workplace communications strategy.  Conversation analysis gives a more accurate picture.

We converse spontaneously, but that doesn’t mean our conversations are random. Actually, most conversations are made up of key building blocks and follow predictable patterns. Understanding conversational elements and how they fit together is the best method for analyzing and improving how we talk, and for avoiding conversational pitfalls that create friction and misunderstandings.

Trying to improve communication strategies at work? Don’t reach for a tired role-play. Recording and analyzing actual interactions will give you far more accurate insight into how your communications can be enhanced. The best way to do this is to create a body of real-time recorded conversations. With your employees’ permission, start recording real-time interactions. Phone conversations or meetings are a great place to start. Once you have enough raw material, you’ll be able to identify conversational problems and strategize solutions.

Check out my related post: How do you spot a lie?

Interesting reads:


  1. Great article! Even in my graduate communication program, I sometimes heard other students tout this statistic without understanding the limitations, and intent, of the Mehrabian study. More comprehensive studies have found that nonverbal elements can comprise up to 70% of a message, but that “up to” is important, because there are a lot of situational factors involved. For example, the ability to correctly interpret nonverbal messages effectively depends a lot on the presence or absence of norming. Thanks for your analysis of this topic.

    Liked by 2 people

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s