First Impressions


Authors: Jean-François Manzoni Jean-Louis Barsoux

For better or worse. a new leader forges relationships in the first few days after taking charge

The first 90 days in a new leadership role are incredibly demanding. What many leaders don’t realize is that the first couple of days may be crucial in developing vital relationships with their subordinates and vice versa. Getting off to a bad start can be difficult to reverse. How you begin sets the tone and establishes the dynamics of key relationships.

When they start in a new leadership role , executives are pulled in multiple directions all at once. The growing literature on leadership transitions highlights numerous priorities, from assessing capabilities and coming to grips with the culture, through to building credibility and managing the network. So much to do, so little time…

Yet, we would argue that among the many priorities identified, the need to lay sound foundations with direct reports deserves special attention because once these relationships start to falter, they consume inordinate amounts of the leader’s time and energy. As one top executive told us, “When you arrive, you first have to set things straight with your direct reports. That will allow you, once they are firing on all cylinders, to free up the time to attend to other constituencies and activities.”

Transition specialists, such as our colleague Mike Watkins, typically talk about the first 90 days as the crucial window.1 But where boss-subordinate relationships are concerned, it is more like the first 30 days – and there is compelling evidence to suggest that it sometimes happens even more quickly.2

Our own research suggests that bosses seriously underestimate how fast they and their direct reports can form negative impressions of one another, and how quickly these impressions trigger vicious performance spirals – with potentially disastrous consequences for the energy and productivity of those concerned, including the team as a whole.3

To avoid such dynamics, incoming leaders have to be conscious of two big labeling traps and the reinforcement mechanisms that give those traps their grip.

Trap #1: The new leader gets pegged

When a new leader comes in, anxiety levels rise. Highly rated executives may worry that the capital they built up with the old boss will not count for much. Those who were previously out of favor may fear that the predecessor’s prejudices will be passed on to the new incumbent. Either way, executives will be concerned and will size up the new

hen they start in a new leadership role, executives boss and his or her intentions so as to anticipate likely threats and opportunities.

The new leader’s early utterances and behavior are closely scanned for patterns, meaning and clues. How does the new boss measure up to each subordinate’s image of an “effective leader”? Does the new boss seem approachable, credible, committed and fair?

A big problem for incoming bosses is that their subordinates often misperceive intentions. When people do not know why their new boss is acting in a particular way, they often deduce that it has to do with the boss’s personality, rather than the constraints imposed by the situation (such as a lack of time, information or resources).

For example, the boss’s failure to consult or inform one or more subordinates about a decision may be enough to trigger a preliminary label like “S/he doesn’t seem very open.” Executives develop these labels to guide their interactions with the boss, to help them anticipate the boss’s likely reactions to events or outcomes, and to protect themselves.

Of course, it is the same the other way around too. Bosses also scrutinize the words and behavior of their individual direct reports – and are just as prone to discern false patterns and to draw false conclusions.

Trap #2: Leaders also label

Keen to establish some momentum, the new leader often has to make early decisions with only a partial understanding of the context. The new boss, therefore, feels compelled to determine quickly whose opinions he or she can trust and who can be counted on for support. The speed with which bosses identify their “lieutenants” is highlighted by a line of research known as leader-member exchange (LMX) theory. The research suggests that most bosses – some studies indicate over 90% – quickly divide subordinates into two clusters: those on whom they can rely for help (the “in-group”) and those they would be less likely to turn to for help (the “out-group”). These distinctions have been shown to emerge within as little as five days of working together.4

Leaders make these early calls using their own rules of thumb about what differentiates higher and lower performers, such as displays of intellect, engagement, cooperation, self-confidence or initiative. Of course, those subordinate attitudes and behaviors are easily misread – and, given the very short time frame, the potential for error is high.

Now, this would not matter if the labels were easily altered. But this is not the case. In studies tracking boss–subordinate relationships over time, the subgroups identified within one month of the new boss’s arrival remained surprisingly stable throughout the course of the year.5 In other words, executives who are misjudged at the outset will find it very hard to change the boss’s perception.

The same is true of a subordinate’s first impressions of the boss. Once the labeling process is activated, two reinforcing mechanisms kick in – one cognitive and one behavioral – which make it unlikely that the labels will be revised.

Labels guide interpretation

The initial impressions that the boss and subordinate develop about each other distort the way they then process information, particularly the way they interpret behaviors and outcomes. The tendency to bend evidence to match our expectations – and to keep our labels intact – is known as attribution bias. It is especially prevalent in organizational contexts, where the complex relationship between actions and consequences is open to divergent interpretations. And it is heightened when people operate under conditions of stress or at some distance from one another.6

Boss Bias

Studies have shown that when in-group subordinates experience a setback, bosses tend to minimize it by blaming it on bad luck or other external factors (like an unreasonable client or lack of support from another part of the organization), whereas for out-group subordinates, it tends to be attributed to lack of effort or low ability.

Conversely, in the event of success, in-group subordinates will be credited for their judgment, competency and efforts, while with out-group subordinates, the boss will tend to favor a contextual explanation (such as “she probably got help” or “the client must have been desperate”).

And it is the same with actions that go beyond the call of duty. When in-group subordinates exhibit such behaviors, the likely explanation is that they are committed team players. Identical displays from out-group subordinates may be dismissed as attempts at impression-management and, therefore, unauthentic.

Subordinate Bias

Subordinates also make skewed attributions. Those whose emerging relationship with the boss is less close are especially prone. Boss decisions that have a negative impact on them may be taken more personally – as evidence of disregard or punishment – even if the boss had little to no degree of freedom. Similarly, errors of omission will be read as errors of commission.

By contrast, subordinates who pick up signals that they are part of the emerging in-group may be more inclined to give the boss the benefit of the doubt on some issues, even when the outcome goes against them. For example, a delay in responding to a suggestion will be understood as evidence of the boss’s heavy workload, rather than a signal of contempt for the idea and/or its sender.

Labels Guide Behaviour

Beyond influencing our reading of others, labels also serve to guide our behavior toward them – and hence their responses toward us.

Boss Behaviour

To head off potential performance problems, the boss increases the attention devoted to the subordinates whose performance is showing signs that they might “need help” – often providing more specific instructions up front, making more forceful suggestions, monitoring results more closely and getting involved at the first sign of trouble. Such measures are meant to help, but they often have the opposite effect, for two reasons:

  • These subordinates find themselves “boxed in” and unable to shine. For example, how can they hope to match the standards of in-group colleagues when they are given routine assignments, fewer resources and little latitude to experiment?
  • The boss’s more controlling approach hits their motivation, by conveying low expectations. Feeling over-monitored and under-appreciated, out-group subordinates come to lose confidence and stop taking risks or proposing ideas, figuring that the boss is not going to appreciate their efforts anyhow.

These findings converge with research into the Pygmalion/ Golem effects, showing that false information about people’s potential unwittingly affects boss behavior – and that significant differences in subordinate performance can show up within as little as one week of the false labels being handed out to the person in charge.

Subordinate Behaviour

Subordinates who have a bad first impression of their boss also behave in ways that tend to provoke self-fulfilling responses. For example, they may be wary or defensive in their initial exchanges with the new boss, making the boss suspicious and inviting close scrutiny.

Similarly, executives who suspect that the new boss is harsh or unfair may discount corrective feedback, viewing it as a personal attack. Of course, when the boss realizes that the feedback has been ignored, he or she may be tempted to turn up the volume and issue even more forceful instructions, thus confirming the subordinate’s negative view of the boss.

Executives who have had a raw deal under the predecessor may even set out to test the new boss, for example, by rehashing past injustices. For the overstretched boss, it may be difficult to react constructively, instantly confirming to the subordinate that the newcomer is no better than the previous boss was.

In other words, just as bosses can drive good subordinates to underperform, subordinates can drive good bosses to behave unreasonably. Whoever triggers the labeling process, chances are it will soon be activated in both directions, producing a vicious spiral that cannot self-correct.

Some of these dynamics may become very painful, very time- and energy-consuming and very debilitating in terms of team spirit. Others will simply remain unsatisfactory, yet bearable.

The converging evidence from both the Pygmalion effect and the LMX research suggests that these suboptimal dynamics are really set off in the very first week(s) of the new relationship. So preventing their occurrence has to be an early priority.

Avoiding false starts

From our research and consulting work with incoming leaders at all levels of the organization, four key recommendations emerge:

1. Frame the relationship

Bosses can choose to be more or less involved in shaping the first impressions of their direct reports. Frequent contact in the formative period of the relationship allows the boss to communicate with subordinates about key priorities, performance measures, time allocation and even expectations on the type and frequency of communication. Bosses can be more explicit about their own style, how they work, the principles they hold dear, what they like and what drives them crazy. This kind of clarity goes a long way toward eliminating some of the confusion about what is expected and/or tolerated.

At the same time, these interactions give bosses an opportunity to get to know their direct reports personally, to develop mutual respect and hence start establishing a difference between the person and the performance.

One way to start communicating these guidelines and values is to develop a “teachable point of view.”7 This is an explanation distilled from and illustrated by the leader’s personal experience. It helps new leaders, particularly in the first few weeks, to seize everyday opportunities to be more explicit about who they are, what they stand for and where they are heading – the aim being to eliminate some of the guessing games about what they really mean or what their hidden agenda might be.

2. Create an open supportive climate

Investing time – the new leader’s scarcest resource – in individuals signals the boss’s commitment to them and starts to build up goodwill. It is that store of goodwill that will help the new leader ride out the initial mistakes and get the benefit of the doubt on ambiguous or painful decisions.

Developing a personal rapport with subordinates reassures them about reporting problems and asking for help. Why? Because they know that they can still be respected as individuals even if their performance falls short of their own and their boss’s expectations.

Making this person–performance distinction also contributes to decreasing subordinate anxiety and defensiveness associated with future feedback from the boss. Critical feedback becomes a negative comment on the subordinate’s performance at a given time and place and under given conditions, but not a commentary on the subordinate as a person.

3. Keep an open mind (and beware of labeling)

The urge to label is deep-seated, so advising leaders not to label would be unrealistic. But leaders can, and in fact must, become more conscious of this labeling process. In particular, bosses need to work hard to resist broad, abstract or generic performance labels, such as “stronger” and “weaker” performer, which are simplistic and therefore poor guides for action.

New leaders need to be especially wary of the performance labels handed out by their predecessors. Past underperformance can have many root causes, such as a lack of confidence, skills, understanding or effort. So new leaders need to keep an open mind about subordinates. This requires some effort, since our natural inclination is to process information in a way that supports our initial observations. To fight this confirming evidence bias, new leaders must override their “default” system. In particular, they need to give due consideration to situational factors when trying to understand behavior or outcomes, especially for supposedly weaker performers.

4. Intervene Early

To keep individual relationships on track, perceived problems need to be tackled as they emerge. Bosses often try to withhold negative feedback at the outset, assuming that it might spoil the development of a working relationship. But provided the boss remains open-minded about the causes of the problem, this need not be a disturbing experience. If a subordinate is doing something that concerns the new boss, that person needs to be told sooner rather than later. While it may not be pleasant, corrective feedback delivered early can be accepted as part of the normal adaptation process.

Delayed intervention only raises the threat and embarrassment attached to the issue. It gives the feedback a punitive edge and reduces the chances that the subordinate will react constructively to the feedback.

A simple habit is to discuss both the data and the interpretation with the person concerned. What prevents bosses from doing this is precisely the assumption that they know what is driving the outcome, be it lack of skill, judgment or effort.

When bosses make the effort to check, they often discover that they have misread the subordinate’s conduct or motives, as well as overlooked mitigating factors. The chance to validate the boss’s assumptions is another key to establishing goodwill with direct reports. It gives them confidence that they will be evaluated fairly.

The team can’t wait

Clearly, these recommendations require new bosses to spend more time and effort talking to their subordinates and getting to know them, as well as processing information and checking assumptions.

Some executives tell us they have more pressing concerns. But to set up productive relationships with all their subordinates, not just some of them, bosses need to devote energy up front, while they – and their subordinates – are still impressionable. Bad dynamics take root fast. Bosses who fail to invest early on can look forward to the prospect of spending steadily more on remedial action, trying to fix or end working relationships that become ever more difficult.

Executives also sometimes retort that processes like labeling and biased attributions are part of life and of “natural selection.” Subordinates who turn out to be poor fits will be eliminated, making way for people who are handpicked by the leader. This argument is not unreasonable. As described above, our research shows that bosses often contribute to subordinate underperformance, but we are certainly not arguing that every relationship is potentially perfect or that the subordinates in place are necessarily the right ones for the situation or the new boss. Removals are sometimes called for. Yet, these removals entail costs, in terms of lost competencies, experience and relationships within the organization. Replacements will need time to get up to speed, and hence, their impact will be delayed. Nor are they guaranteed to work out exactly as expected.

We are not arguing that leaders should invest indefinitely to make every relationship work. We are encouraging bosses to invest early, in order to increase the likelihood that the potential of the subordinate and the relationship will be maximized and that incompatibilities will be identified early, before they have had a chance to create too much damage. And we are drawing leaders’ attention to the fact that when they perform this assessment, they must remain mindful of their (human but nevertheless potentially dysfunctional) cognitive biases.

Breaking a negative dynamic once it has gathered momentum is more complicated. It requires both parties to set aside their mutual labels and biases, to look honestly at what each of them is doing that disturbs the other party, and to come up with a realistic remedy. This is a complex and sometimes painful process, with an uncertain outcome, which we explore in detail elsewhere.8 So, the dynamic is much harder to interrupt than it is to prevent and start on the right foot.

Questions to Ask

Your first week in your new role just passed. What have you accomplished so far? What do you have to do next? Think about a few questions:

  • How did you introduce yourself to the team? Did you talk about your work style? Were you transparent? Were you curious? Do they understand your expectations and what you will be judging them on? Were you clear about the two or three critical dimensions on which you expect each of them to excel?
  • How many of your direct reports have you spoken to face to face? What do you know of their strengths and weaknesses and what makes them tick? A good question for you, as boss, would be: “What do you hope for from me?” It also paves the way for you to be explicit about: “And now let me tell you what I hope for from you…”
  • Are some labels starting to appear in your head? How will you make sure they don’t become self-fulfilling? Can you talk to people who havea different view of the individual to see what you may be missing?
  • If someone has done something that concerns or irritates you, have you spoken to them about it yet?
  • When giving feedback, did you start by asking, rather than telling? Did you first check that what you thought you saw or heard indeed happened and, if so, why it happened – before issuing warnings?
  • Have you expressly invited your direct reports to help build the working relationship and let you know when your actions are not helpful for them?

  1. Watkins, M. The First 90 Days: Critical Success Strategies for New Leaders at All Levels. Boston, MA: Harvard Business School Press, 2003. Also, Ciampa, D. and M. Watkins Right from the Start: Taking Charge in a New Leadership Role. Boston, MA: Harvard Business School Press, 2005.
  2. Liden, R. C., S. J. Wayne and D. Stilwell. “A Longitudinal Study on the Early Development of Leader-Member Exchanges.” Journal of Applied Psychology, Vol. 78, Iss. 4, 1993: 662–674.
  3. Manzoni, J.-F., and J.-L. Barsoux. The Set-Up-To-Fail Syndrome. Boston, MA: Harvard Business School Press, 2007.
  4. See respectively, Liden, R. and G. Graen. “Generalizability of the Vertical Dyad Linkage Model of Leadership.” Academy of Management Journal, Vol. 23, Iss. 3, 1980: 451-465; and Liden, R. C., S. J. Wayne and D. Stilwell (1993) “A Longitudinal Study on the Early Development of Leader-Member Exchanges.” Journal of Applied Psychology, Vol. 78, Iss. 4, 1993: 662–674.
  5. Dansereau, F., G. Graen and W. J. Haga. “A Vertical Dyad Linkage Approach to Leadership within Formal Organizations: A Longitudinal Investigation of the Role Making Process.” Organizational Behavior and Human Performance, Vol. 13, 1975: 46–78.
  6. Cramton, C. D. “Finding Common Ground in Dispersed Collaboration.” Organizational Dynamics. Vol. 30, Iss. 4, 2002: 356– 367.
  7. Barnett, C. and N. Tichy. “How New Leaders Learn to Take Charge.” Organizational Dynamics, Vol. 29, Iss. 1, 2000: 16–32.
  8. See Chapter 8 “Cracking the Syndrome” (pages 161–195) in Manzoni, J.-F. and J.-L. Barsoux. The Set-Up-To-Fail Syndrome. Boston, MA: Harvard Business School Press, 2007.

Download the original article