Managers hate to give negative feedback, and many defeat the purpose of a feedback discussion with employees by setting up a confrontation. So says Jean-Francois Manzoni in the September issue of the Harvard Business Review. Manzoni, the associate professor of management at Insead in Fontainebleau, France, and the director of the Insead-PwC Research Initiative on High-Performance Organizations, says taking a different approach to giving feedback can make those discussions -- and your employees -- more productive. He discussed his ideas with Kathleen Melymuka.
Q: What is the most common mistake IT managers make in giving corrective feedback to their employees? IT managers' problem with feedback starts with the way they frame -- mentally construct -- the situation. Feedback is something that bosses have and subordinates get. The implicit complement is that this feedback is right, of course.
Q: How might this play out?
"Bill's performance is not up to my expectations. I know why: Bill has the following shortcomings/skill or character deficiency. I want to tell Bill about this, but he may not like hearing what I have to say. He may be hurt, and he may try to hurt me in return. Also, if Bill refuses the feedback and pushes back on me, things may escalate and worsen the current situation. So I hope Bill will accept the feedback and spare us an unpleasant moment."
Q: What's wrong with this approach? This framing is narrow because it excludes some potentially interesting issues and questions such as: Am I right? What's the evidence on which my assessment is based? Could I be missing part of the picture? How much is it really about Bill, vs. about the situation Bill is operating in? And what about me? Could I be contributing to the situation, and if so, how?
In addition to being narrow, this framing is binary in that there are only two possible outcomes: The session is a success if Bill accepts the feedback; otherwise, it's a failure.
Q: Why is this mistake particularly likely to happen in an IT environment? IT environments present three characteristics that make it particularly hard for bosses to approach feedback productively. First, the success of IT projects tends to be very dependent on the actions of many parties outside IT, particularly IT users. When looking at results, it is hence difficult to untangle the exact contribution of the IT staff from the impact of the conditions they were operating in. Second, IT is often a bad-news-driven environment. Most of the feedback IT receives from the organization tends to be negative. Last, several aspects of IT involve work that is intangible and largely invisible until completion. Assessing progress intelligently is difficult.
Q: You say the situation worsens if the framing remains frozen during the discussion. What does that mean? Bosses rarely revise their position during the encounter, even when the subordinate brings up potentially relevant information. That lack of flexibility prevents a more effective handling of the situation and can lead to an escalation in tone and/or content of the discussion.
Q: Why don't IT managers just revise their restrictive framing midstream, when they see it isn't working? First, they are not conscious that they framed the encounter in a narrow and binary way. It's hard to consciously revise a mental construction that we don't know we have. Second, the more intense the discussion, the more energies are diverted to keeping oneself and the discussion under control, and the smaller the bandwidth available to process fully the information we receive.
Q:There's a different approach you call "easing in." What's that? Easing in is an attempt to avoid a collision with the other party. So rather than telling you what I have in mind, I am going to ask you a series of questions which, if you answer them correctly, will lead you to the "right" conclusion -- [the one] I have already made. A typical easing-in question is: "Don't you think that. . . ?" It really means: "I think this way, and I really hope you'll agree."
Q: You say easing in is a gamble that often doesn't pay off. Why? Easing in is successful only when the subordinate gives you the "right" answers and does not realize he is being manipulated.
Q: How does easing in go wrong? If the subordinate fails to give the "right" answers, the boss must either fold and try again later, or else make his point more explicit. And the subordinate [may] realize somewhere along the way that this is not really a discussion: He is being led carefully to a prepackaged conclusion. Most people don't like feeling manipulated.
Q: So aside from failing to achieve their purpose, these approaches can damage relationships? The frontal delivery of [negative or corrective] feedback often leads to escalations, resulting in one or both parties saying things they didn't really want to say, and/or in one or both parties giving up and pretending to agree. The easing-in approach can lead to an escalation if the subordinate does not "play along," and it can lead to the subordinate pretending to comply if he realizes the boss has already made up her mind. None of these outcomes is very constructive.
Q: What are the conditions that make critical feedback more acceptable to employees? Research suggests that feedback receivers are more likely to accept and act on feedback when they feel that the feedback source has good intentions toward him/her; developed the feedback fairly, which includes collecting all relevant information, allowing the receiver to provide clarifications/explanations and applying consistent standards; and communicated the feedback fairly by showing openness toward and support for the receiver.
Q: Can you give me an example of the right and wrong way to open a feedback discussion? Bosses should approach feedback with a mental framing along the following lines: "I am not happy with Bill's present performance, nor with our relationship. I think I understand where Bill's problems are coming from. But I could be wrong. Bill probably feels the malaise and wants this job and our relationship to work at least as much as I do. So we both want the same thing but somehow we're not getting it right. Let's sit down and discuss why this is the case and how I can help."
This framing is not binary -- there's no clear pass/fail criterion. It is broad and, hence, flexible. It is a good basis for a real discussion. If you think this way, the words will come out right.