Trust on credit
Some time ago a manager I know came to me for advice. I am telling this story with her consent, but without her name, because that is what she asked for. Her organization had just finished a round of anonymous surveys, and it turned out that a large part of the team had a problem with what it was like to work with her. On top of the results came the conversations her boss had held with team members, and their tone turned out to be the same. So he gathered it all together and delivered the devastating news to her personally.
A few days later she came to me asking what to do with it. Showing people the areas they cannot see themselves is a regular part of my work, both in my own company and with the leaders I advise. So we sat down to talk about her difficult feedback.
There is no hiding that all of it was churning inside her. The hardest thing for her to believe was that the picture from the surveys was real, because on a daily basis nobody ever told her anything like it. People she had worked with for years smiled, cooperated and raised no objections.
I told her honestly that in her place I would also want to go to people, hear everything directly from them and be able to ask for details. Even so, I advised her against it. I knew her, and I knew the situation well enough to know this was not the moment for that kind of round. People give us honest feedback straight to our face only when we have credit of trust with them, and her credit had just melted away. It had to be rebuilt first. So together we sketched a plan of action, where to start and what to leave alone for now.
The catch is that this was not the first situation like this I had seen in my life, so I was almost certain she would not listen to me anyway. Difficult feedback has this to it that it practically begs to be verified at the source.
A walk around the company
Some time after our conversation she did exactly what I had warned her against. She went around most of the people in her company, asking them, face to face, what problem they had with her. I learned about all of it when we met the next time.
Almost everyone reacted the same way, backing off and assuring her they had no problem with her at all. People who had described their difficult experiences in the anonymous surveys, suddenly asked point blank by the very person it all concerned, denied everything (I imagine some of them in plain panic).
She told me about it with pride. Everyone had been wrong, the surveys were lying, and the reality she moved in every day looked completely different. She held a full set of assurances that everything was fine, and I find it hard to blame her for believing them. That other version, the one from the surveys, did not have a single face after all.
I told her plainly that it had been a mistake, and that her reality could be entirely true for her while the truth still looked different. That conversation left me with a lot of sadness and helplessness. Sometimes, when I try to point people in a direction or show them a path so they do not have to learn from their own mistakes, I watch it simply bounce off them. They do not take in what I am trying to show them, and later, probably without even knowing it, they sabotage their own next steps.
If this were a one off case, I probably would not be writing this piece. But I see the same pattern every now and then in more people, in different companies, and even in my own. The people and the circumstances change, yet the story unfolds in a very similar way. Difficult feedback, disbelief, checking at the source on your own, and a chorus of denials at the end.
The simplest conclusion would be that she worked among people who lied straight to her face. Except I have heard this same ending far too many times to believe that explanation.
No one lied out of malice
After this story, to show its wider context rather than just personal experience, I started looking for an answer to the question of why people do not tell the truth at work even when asked directly, and why almost all of them, one after another, chose denial. It turns out psychology has known this phenomenon for over fifty years as the MUM effect, from the English phrase “keeping mum”, that is, holding your tongue. The name itself comes from the experiment that started this line of research.
A participant was waiting for an experiment to begin and in the meantime learned that a phone message had arrived for another subject who was about to walk in. That person was supposed to urgently call home, because something very bad had happened. When the addressee appeared, participants usually passed on the practical part of the message, “call home”, far more rarely adding that it was about bad news.
Sidney Rosen and Abraham Tesser described this tendency in 1970 in “Sociometry”. In short, when we have something unpleasant to pass on to someone, we delay it, soften it or stay silent altogether. And we do it even when the bad news is not our fault, when it simply fell to us to deliver it.
The first explanation that came to my mind is that we are simply sparing each other the discomfort.
Seventeen years later Charles Bond and E. L. Anderson tested it in an experiment published in 1987 in the “Journal of Experimental Social Psychology”. In the study, a participant had to give another person their intelligence test result, either excellent or very poor, according to a key prepared by the researchers. The result was fixed in advance. The messenger had nothing to do with it.
When the participant believed the recipient could see them, they delivered news of failure twice as slowly as news of success. When they believed they remained invisible, they spoke about failure just as quickly as about good news. The authors saw concern for one’s own image in this, though a single small study does not allow us to treat it as the only explanation.
Half a century after the first experiment, Abi-Esber and her team checked in a series of studies published in 2022, on nearly two thousand people in total, how often we tell others about things they could fix on the spot. Over three lunches, one of the researchers ran short surveys with a red line, a chocolate smudge or smeared pink lipstick on her face. She talked to people in a way that made them look at her.
Out of 155 people who later confirmed they had noticed the problem, four told her about it. The remaining 151 looked, talked and stayed silent. In their explanations people focused mostly on their own awkwardness or on possibly embarrassing the researcher. Hardly anyone thought in that moment about the benefit she would have gained if someone had actually pointed it out.
Meanwhile at work, the price of honesty can be much higher than a few seconds of awkwardness. Frances Milliken, Elizabeth Morrison and Patricia Hewlin, in interviews with forty employees of various companies published in 2003, heard from as many as 85% of them that at least once they had not raised an issue with their superior that they themselves considered important. They were held back by fear of being labeled difficult, of damaging relationships, of retaliation, or by the belief that speaking up would change nothing anyway.
When I put those numbers next to what happened in my friend’s company, I see that the people she questioned did nothing unusual. Most of us would probably behave the same way if someone stood in front of us and asked about themselves. The answer “I have no problem with you” ends an awkward moment in a few seconds and risks nothing, while an honest one opens a conflict and puts at risk a relationship with a person you will sit down to work with again tomorrow. None of them owed her that kind of courage, especially since the credit of trust I wrote about earlier was something she had not managed to rebuild with any of them.
A separate matter is the question itself that she asked them, because it also gave them no chance at honesty.
A question that answers itself
To be fair to her, that walk around the company took real courage. I do not know many people myself who would go up to a colleague with the question “listen, do you have a problem with me?”, because we are afraid of what we might hear.
Except when I put myself in the position of the person being asked, I see that I really have two options. I can say “no, everything is fine”, which closes the subject immediately and costs me nothing. Or I can say “yes, I do”, only then I have to justify it, give examples and enter a dispute with a person standing a meter away from me, weighing how it will affect our everyday cooperation. On top of that, in this case the person asking was someone with whom difficult experiences had already piled up. A question like this, though it sounds like the height of openness, suggests the safe answer all by itself.
Nobody, as far as I could find, has directly studied the question “do you have a problem with me?” asked face to face in a company. A lot is known, however, about how strongly the channel of an answer changes its content. Interestingly, in a study published in 2000 in the “Journal of Applied Social Psychology”, Ong and Weiss checked how many students would admit to cheating.
In the first stage, students entered a small room one at a time. On the monitor waited a ten minute test of twenty words. Five were easy, the rest very hard, and the last three the researchers had simply made up. For at least seventeen correct answers you could get 25 dollars.
About a meter from the monitor, among six books, lay a small dictionary. The students had been told beforehand that they were not allowed to use any materials. Fifty five out of seventy nine people reached for the dictionary anyway. The researchers knew this thanks to its position and a bookmark left there on purpose.
At least three weeks later, a second researcher came to class with a seemingly unrelated survey. Among other things, he asked about using forbidden materials during tests. In one version, the student wrote their full name and social security number and handed the sheet directly to the researcher. In the other, they gave no name, wrote only the last five digits of the number and dropped the form into a cardboard box.
In the confidential but not anonymous survey, 25% of those who had actually cheated admitted it, and with greater anonymity it was already 74%. It was enough that the person answering stopped being identifiable.
After all this, it is hard for me not to think that the amount of truth in an answer depends above all on how the person giving it feels. That is why I advised her to rebuild trust first, because trust in a team is the precondition of an honest conversation, and she wanted to earn it with that one round of questions.
If she had come back from that walk empty handed, it would have been half the trouble. But she brought back answers, and it was the answers that did the most damage.
A blind spot painted over
Each of us has a point in the eye where we see nothing. Where the optic nerve leaves the eyeball there are no light receptors, so when looking with one eye there should be a hole in the field of vision (you can check it with a simple test using a dot drawn on a piece of paper). We do not see it, because the brain paints that spot over with whatever surrounds it. Instead of emptiness we get an image constructed from the surroundings, one that is not really there.
It is similar with knowledge about ourselves. The things we cannot see about ourselves do not look like holes, because the mind fills them with the most comfortable available version. That manager went looking for her blind spot, which in itself was a good decision. But she came back with material that could at best paint it over, a series of answers saying there was nothing there.
Peter Ditto and David Lopez showed in a series of experiments published in 1992 that we accept information consistent with what we want to hear about ourselves almost without scrutiny, while we set much higher evidential standards for the uncomfortable kind.
The researchers invented a fictional deficiency of the enzyme TAA, which supposedly increased the risk of pancreatic disease. Participants dipped a yellow strip of paper in their saliva and watched whether it changed color. The strip was ordinary construction paper and always stayed yellow. Half of the people were told that no change meant health. For the other half, the same lack of change was supposed to be bad news.
When the yellow strip meant a favorable result, participants finished the test after about 76 seconds. With an unfavorable one, they waited 105 seconds on average. More than half repeated the attempt, while after good news roughly one in five people did. Some dipped the strip again, others took a new saliva sample, shook the paper or blew on it. One participant dipped it twelve times.
The authors of the study called this behavior motivated skepticism. In this story, the denials heard face to face went in instantly, while the results of the anonymous surveys were the thing that had to be checked and ultimately declared a mistake. It comes through even more strongly in the study that Oliver Sheldon, David Dunning and Daniel Ames published in 2014 in the “Journal of Applied Psychology”. They examined how people react to the results of an emotional intelligence test. Those who did worst inflated their self assessment the most, and after seeing their results they more often questioned the validity of the test itself and were the least willing to invest in developing that skill. So it seems that the people who most lack people skills find it hardest to accept information about that lack. I am far from diagnosing her with these studies, especially since I can defend myself against uncomfortable information just as efficiently, only usually less spectacularly.
My friend went on believing for a long time that there was no problem. She had heard it, after all, from most of her company with her own ears, and everything that said otherwise, the surveys, her boss’s conversations with the team and my advice, she set aside as not worth believing. What remained was anger, and after hard news anger rarely aims at the content. Most often it hits the person who brought it.
The lesser evil of the messenger
Her boss collected most of that anger. He was the one who laid the survey results in front of her, and he was the one who summarized his conversations with the team. From her account it was clear that he was the hardest one to forgive. As if the messenger were responsible for the content of the message he brought.
Leslie John, Hayley Blunden and Heidi Liu, in one of eleven experiments described in 2019 in the “Journal of Experimental Psychology: General”, showed that we judge an innocent messenger of bad news as less trustworthy and less likable. In the study, an assistant drew a card from a hat and, without reading it, passed it to another person. Only that person informed the participant whether they had won an extra two dollars. The messenger did not choose the number, did not draw it and had no influence on the money.
Despite that, people who lost rated him as less likable than participants who won. In another experiment, in which subjects imagined a visit to a clinic, after a bad biopsy result the rating of the nurse delivering the diagnosis dropped, but not of the person who only scheduled the next appointment. So the change did not apply to everyone encountered along the way. It touched the one who became the face of the bad news. The researchers call it shooting the messenger. The anger at her boss was predictable, even though he was the least at fault here.
I have carried someone else’s difficult feedback more than once myself, so I know very well the situation that man found himself in. The intermediary has no good way out here. He can point to who said what, and burn the trust of people who spoke up precisely because they were meant to stay invisible. Or he can speak in general terms, about results and patterns, accepting that such a message is much easier to dismiss. Stephen Rains described in 2007 in “Communication Research” the discounting of an anonymous source, meaning that we judge information from a sender who cannot be identified as less credible and take it less seriously. An anonymous survey is such a source by definition. Since nobody knows who wrote what, the whole thing could be brushed off as measurement error or someone’s malice.
In the shooting the messenger experiments, the researchers also noticed that dislike of the sender decreases when they openly show that they bring the message out of care for the recipient. I told my friend about this when we came back to the story. I explained that her boss had done a difficult and thankless thing, and that the openness she was looking for in her walk around the company is built through everyday engagement, not through one round of questions.
Back then it had not sunk in yet. Sadly, it ended with her eventually being let go. Today she is much richer for those experiences, and in the company where she works now she has become an ambassador of 360 surveys herself, exactly the kind of feedback she once rejected.
I could close this story with a moral about her. Except I have a blind spot too, and I read these same studies against myself as well.
How I know what I do not know about myself
For years, usually twice a year, and in the worst case once, I have been asking people in my company for an anonymous survey about me. I ask about my working style, about communication, about the level of trust in me and about how it all looks from their perspective. It is mostly my closest collaborators who answer, but the whole company has the chance to speak up. It is the same kind of knowledge she received in her surveys, except I ask for it.
I did not always do this. In my first businesses, when we were a team of a few people, I simply asked people directly whether they would share their observations with me. Looking back, I can see I did it far too rarely. I was a young man and the need had not yet become second nature.
Today I treat feedback like this a bit like a gift, because it is the only access to my own blind spot that I have. I will not learn these things on my own. Most of what I receive today is admittedly positive, but what I am really waiting for in those surveys is the less pleasant part, because only that gives me a chance to still change and improve something. I have the impression that it is precisely those difficult things, together with the attempt to work through them honestly, that made today’s balance more good than bad. I do not want this to sound as if I were waiting for someone to torment me. I mean constructive remarks, like the time when, under too much overload, people pointed out details that were escaping me. That helped me rethink when I should be making decisions and when I should not. Today there are days when I do not enter the decision making process at all, because I know a decision made in that state would be wrong. And I would not have gotten there without self reflection, deepening awareness and precisely this collecting of feedback.
This truth about myself does not come to me for free, though, even though the surveys are anonymous. I have to ask for it regularly. People need to trust me enough to want to write it, and they need to know me well enough for their picture not to be blurred. What good is someone’s opinion of me if they do not know how I work and how I make decisions. Besides, when the blurred answers start piling up, that too is information for me, that I am too far away from people.
My experience with these surveys matches well what is known about asking for honest feedback. Susan Ashford and Anne Tsui studied 387 managers in 1991 in the “Academy of Management Journal”, and it turned out that those who actively sought criticism were rated by superiors, subordinates and coworkers as more effective, while those who sought mostly praise came out worse in those ratings. That is of course correlation, not proof, but the direction is hard to ignore. Hayley Blunden, the same researcher who worked on shooting the messenger, showed with her team in “Management Science” that when we ask for advice instead of feedback, we get a third more areas for improvement pointed out and over half more concrete suggestions on how to improve. In a request for feedback people hear a call to judge what has already been, while a request for advice directs their attention to what can be done better. At least that is how the authors of the study explain it.
If I were to offer anything from this story, it would be two questions rather than a recipe. Does the person I want to ask about myself have reasons to believe that honesty will cost them nothing with me? And am I asking them for advice about the future, or for a verdict on the past? Without a good answer to the first question, the second no longer matters much.
And finally, I ask myself the same question I am leaving you with. Who today trusts you enough to show you your blind spot, painted over so carefully that it looks like a healthy field of vision? And when was the last time you heard something about yourself that was not a polite “everything is fine”?
