Share Podcast
The Persuasive Power of Uncertainty
Zakary Tormala, associate professor of marketing at Stanford’s Graduate School of Business.
- Subscribe:
- Apple Podcasts
- Google Podcasts
- Spotify
- RSS
Featured Guest: Zakary Tormala, associate professor of marketing at Stanford’s Graduate School of Business.
SCOTT BERINATO: Welcome to the HBR IdeaCast from Harvard Business Review. I’m Scott Berinato. We’re joined today by Zak Tormala of Stanford’s Graduate School of Business, whose new research suggests that experts who are less confident about their opinions are surprisingly more persuasive and compelling. Thanks for joining us, Zak.
ZAK TORMALA: Thanks for having me.
SCOTT BERINATO: Great, so I want to talk about this really surprising research of yours about how uncertain experts tend to be more persuasive. And tell me how you came to this finding.
ZAK TORMALA: Well, I think when we started this research, we had the same hypothesis that a lot of people would have, that being more certain would be generally better when you’re trying to persuade somebody or convince somebody to do something. And so, we were originally thinking expressing confidence or certainty might be a solution for people who lack credibility to gain influence.
And so, we did some studies to look into that possibility, and found that, in fact, under some conditions it was having the exact opposite effect, that expressing certainty was backfiring rather than giving you a persuasive advantage.
SCOTT BERINATO: So tell me how you went about testing this. What did you actually do with your test subjects?
ZAK TORMALA: We ran a variety of tests that all sort of assumed a similar form in terms of the structure of the studies. And to give you an example that’s pretty representative of what we wrote up in the paper, we told our participants to read a review for a restaurant, a new restaurant that they hadn’t been to. It was actually a fictitious restaurant, but they were led to believe that it was a new one in the area that they conceivably could go to.
And we had them read a review, and the review for the restaurant was, in most of our studies, quite favorable and also compelling, talking about things like the ambiance and the quality of the food and the service and the like. And what we manipulated was who gave that review.
So in some cases, participants read a review from an expert, a restaurant critic, and in other cases they read a review from a non-expert, just some local person who had a blog. And we manipulated how much certainty was expressed.
So, in each case, the person always gave the restaurant a positive review, but sometimes they said something along the lines of, having been there myself, I’m highly certain of this. I gave it a four out of five. Or in the other case, having been there only once, I’m not so certain, but I would give it a four out of five.
And so, we just had sort of varied some of those key variables, and then measured participants’ responses. For instance, how good they thought the restaurant would be, how much they would pay for a meal there, that kind of thing.
SCOTT BERINATO: And so, what you found was that people were willing to pay more to go to the restaurant when the expert reviewed it and was uncertain than when the expert reviewed it and was certain. Is that correct?
ZAK TORMALA: That is correct. That was the surprising finding of the paper, or of our original studies, anyway, that turned into a paper exploring that effect. But we found that, in general, experts who expressed uncertainty about their recommendations had more persuasive impact.
And so, in one study it did manifest in terms of greater willingness to pay. So people who read a favorable but uncertain review from an expert were willing to pay more to try a meal at that restaurant than people who read a favorable and certain review from an expert.
On the non-expert side, though, we flipped that effect. So what we found there was that when participants read a review from a non-expert, more certainty was better. In that case, more certainty did lead to greater willingness to pay. Generally, more favorable reactions overall.
SCOTT BERINATO: So on every variable, this seems like the exact opposite effect that I would expect. So can you explain what do you suspect is going on here?
ZAK TORMALA: Yeah, it is surprising and it’s counterintuitive. But once you sort of think through the logic, it does make sense and it does fit with prior theorizing in psychology and other fields.
And so, basically, what we think is going on, and what we have evidence for, is that an expert who expresses uncertainty is surprising somehow. So people expect experts to be confident in their opinions. If this is a restaurant critic, the person should be highly certain about his assessment of a restaurant.
And so, it’s surprising when the person who’s an expert expresses some hint of hesitation or uncertainty or doubt. And that surprise grabs attention and draws people in. And so, it’s the sort of tension between the person’s expertise and confessed uncertainty that leads people to sort of feel something is amiss here, let me pay more attention and make sense of this.
And so, once they’re paying more attention, they’re sort of more open to influence. They’re being drawn into the message. They’re reading more carefully, and then assuming that the message is compelling. That leads to more persuasion.
SCOTT BERINATO: And on the flip side, if you’re an amateur or a non-expert, you probably expect them to not know their stuff very well. A so, if they express this real deep confidence in their experience, then you’re going to perk up and pay attention.
ZAK TORMALA: That’s right. That’s exactly what we think is happening. So it’s sort of the same core logic in each case, but just different combinations of these variables.
SCOTT BERINATO: Here’s what concerns me about this, Zak, and I immediately start thinking about expert witnesses or eyewitness testimony, and if it’s true that the uncertain expert is more influential, then what do we make of experts on the stand expressing their certainty in situations where somebody’s maybe on trial for a major crime?
ZAK TORMALA: Yeah, I think that’s actually a really good and really important question. We haven’t tested this across domains, and so it’s hard to speak definitively about that, but there is research in eyewitness testimony and a couple other domains suggesting that, in general, being more confident or more certain is more compelling, so a jury is more convinced by testimony that’s given with high confidence. And presumably, that’s true when it comes from an expert, too.
And so, we think– this is our speculation– but we think one of the key differences here is that eyewitness testimony is given in a context in which there often is an objective truth, and so there is a right answer, that the thing did or didn’t happen that’s being testified about. And it could be that, in those kinds of settings, confidence does correspond with accuracy, or at least it’s perceived to correspond with accuracy.
Whereas, in a subjective domain, like restaurant opinions, which are a matter of sort of taste and perception and that kind of thing, the psychology is a little bit different. And so, whether there actually is a right or wrong answer, or a truth to be uncovered versus just a matter of opinion, where there’s different sort of subjective experience when you’re listening to a person express high or low certainty.
SCOTT BERINATO: Now, you did twist one other variable, which I thought was fun, which was you had the non-experts talk about other variables besides the quality of the restaurant. They might talk about the decor of the restaurant or what the menus look like. And tell us what the effect was there.
ZAK TORMALA: Yeah, that sometimes paying more attention leads to less persuasion if what you’re paying attention to is not compelling or is irrelevant. So what we did in the study that you’re asking about is manipulated the quality of the arguments contained in the review.
And so, in some cases, the review talked about the food, the ambiance, the service. These are the traditional markers of a good restaurant. It has good scores on these dimensions.
In other cases, participants read a review that talked about things like the color of the menu, or the font used to print the items, or the conversation that was had at the table, things that are all fine and dandy but not really of central relevance to what makes a restaurant good. And so, what we found was that when the expert expressed uncertainty, that did lead to more persuasion when the review was compelling. It talked about the right stuff.
But when the review talked about irrelevant or less compelling features, that same combination, where the expert was uncertain, didn’t lead to more persuasion. And in fact, it tended to backfire. There was a non-significant tendency for that to backfire.
So getting people to pay attention is good if what you’re telling them is compelling. But if what you’re telling them is not compelling, getting them to pay attention could backfire. That’s sort of the take-home point of that finding.
SCOTT BERINATO: Of course, we’re HBR, so we have to ask the business question, connect this to the business. Let’s change the restaurant critic to a CEO who obviously is an expert on his business, and he stands up in front of the company and expresses uncertainty about the business. Is that going to be a more compelling and more persuasive CEO than one who stands up with all the confidence in the world about what he’s saying?
ZAK TORMALA: Our core prediction or core finding is really about what grabs attention and gets people to process what you’re saying. And so, we do think a CEO should be an expert about the company, presumably. And if this CEO gets up and expresses uncertainty about something, that that might make people perk up and pay more attention, just like we find with restaurants.
And then whether that translates into persuasion or not would depend on what the CEO is saying, just as I mentioned with the other study. And so, generally speaking, I think it should apply there, where if the CEO is trying to figure out, how can I get people to pay a little bit more attention to what I’m going to say today at lunch?, or something like that. One technique might be to express a little bit of doubt or uncertainty at the opening, and that might surprise people and suck them in, and they’ll pay more attention.
And then as long as what you’re saying after that is compelling, it should be more persuasive. But the other thing to remember is that there are multiple goals in persuasion contexts, and one is to make the message itself compelling or have more impact. Another is just to give people confidence or make them feel comfortable.
And so, although an expert who is highly certain might not be surprising, it might not get the same kind of attention as one whose uncertain, that person might be liked, people might like the person who expresses certainty, because that person feels like they’re more charismatic somehow or something else. And that can generate confidence and support and that kind of thing as well.
So our recommendation wouldn’t be that a CEO should always express uncertainty. It would be more dependent on what the CEO is trying to accomplish. So if I want people to pay attention to me today, I might express a little bit of uncertainty at the outset. If I want people to feel comfortable and feel like they have a strong leader who sees a path forward and rally the troops and get everybody fired up, then maybe being highly certain is better, because it’ll sort of make me feel more charismatic somehow. So it’d sort of depend on the goals of the situation.
SCOTT BERINATO: I love this whole concept of certainty and uncertainty. And I know there’s one other sort of interesting area we talked about, which was related but not exactly the same, which is the value of potential over the value of results. And you want to talk a little bit about that research?
ZAK TORMALA: Yeah, sure. So if you evaluate a person who has potential to accomplish something great, that might be more compelling than evaluating somebody who actually has accomplished something great, because you’re paying more attention. It’s more interesting or engaging somehow when it’s potential.
So in one common example, where we feel like we’ve seen this anecdotally, and we’ve also studied examples like this, we ask people to evaluate a graduating Ph.D student who has potential. Like, letter writers, for instance, described this person as having potential to publish 10 top-tier journal articles in his or her first three years as an assistant professor, versus evaluating somebody who’s at the three-year mark as an assistant professor and has published 10 top-tier journal articles.
So the people are exactly the same in terms of the level of performance that’s being considered. But the person either has achieved it or has potential to achieve it. And so, intuitively, the person who has mere potential to achieve it seems, in some ways, less compelling. The person who actually has achieved what the other person merely has potential to achieve you’d think might be slightly more compelling.
In fact, we find the opposite kind of effect, where somebody who has potential to be great is preferred over somebody who actually is great.
SCOTT BERINATO: So there’s this whole world of certainty research in what’s compelling. And I guess we want to hear some of the latest work. I know you’re working on some new things. What are some of the latest projects you’re working on that you can give us a sneak peek into? What are you looking at now in terms of certainty research?
ZAK TORMALA: We’ve been trying to look back at some variables that are well studied in the persuasion domain, to try to understand, are there cases where what we normally think is less persuasive can become more persuasive? And so, the uncertainty stuff is an example of that, where being uncertain is surprisingly more powerful sometimes.
In a recent project with Taly Reich, one of our Ph.D students here at Stanford, we’ve been looking into the possibility that giving conflicting advice, which also sort of might correspond with being less certain, but giving conflicting advice is sometimes more persuasive than giving straightforward and consistent advice.
So, saying one thing, you know, don’t take that class today, and then tomorrow saying, in fact, on second thought, I think you should take that class, sometimes gets people to be more likely to take that class than if you just said take it the whole time.
And so, we think it’s sort of related, where a change of heart or some inconsistency of the advice– although, we normally think that’s a bad thing. That’s flip-flopping and that’s not compelling. Sometimes, because it suggests that things aren’t quite as certain as they might normally be, I should pay more attention and think more about this.
And again, that can sometimes lead to more influence or more persuasion. So that’s one domain where we’ve been looking at sort of related uncertainty-type effects, finding that being a little bit less consistent in your advice-giving is sometimes more powerful.
SCOTT BERINATO: Zak, thanks again for joining us.
ZAK TORMALA: My pleasure, Scott. Thanks for having me.
SCOTT BERINATO: That was Zak Tormala of Stanford’s Graduate School of Business. For more on his research, you can pick up the March 2011 issue of HBR, or visit hbr.org.