A project manager holds two jobs at once, and they are different jobs. The first is managing the work, which means scheduling it, funding it, monitoring it and planning for what to do when something goes wrong. The second is leading the people who do the work, which means getting them to want the outcome so much that they make good decisions when nobody is watching. A project manager who manages the work and leaves the team unled ends up with an accurate schedule and a passive team.
The two jobs are easy to confuse because the same person does both, usually in the same meeting. The sections below separate them, describe the leadership styles a project manager can choose between and explain how to lead people who report to someone else.
How led work differs from managed work
The test that separates the two jobs is what happens to the work when the project manager is out of the room. Work that has only been managed stops or drifts, because the coordination lived in one person and that person is absent. Work that has been led carries on, because the team understands the outcome well enough to make the small decisions the plan never anticipated. Every project produces far more of these small decisions than any plan can hold, which is why a project that is only managed falls behind the moment its manager looks away.
The two factors that decide the leadership style
Leading is what keeps the work moving, and how to lead depends on who is being led. Situational leadership theory states that the right style depends upon the person in front of the leader and the task they have been given, and does not depend upon the natural style of the leader. Two factors govern the choice. Competence is the person's ability to do this particular task, judged on this task alone and on nothing they have done elsewhere. Confidence is their belief that they can do it and their commitment to doing it. The two are independent of each other, and this independence is what leaders most often overlook. An engineer who has done a task ten times may still doubt the eleventh attempt because the tenth went badly. A new employee may be eager and entirely wrong.
Since both factors are measured against a task, the assessment is of the task in front of the person. This means that the same developer needs close supervision on a regulatory reporting change and no supervision at all on the service they wrote themselves. Fixing a style to a job title is the common mistake.
The four leadership styles
Competence and confidence each run from low to high, and the four combinations give the four styles.
Competence decides how much of the method the leader supplies. Confidence decides whether the leader has to stay in the room while that method is used. The same person occupies different cells for different tasks within the same week.
| Style | Competence on this task | Confidence on this task | What the leader does |
|---|---|---|---|
| Directing | Low | Anywhere, often high | Sets the goal, specifies the method, checks the work at short intervals and corrects it in the moment |
| Coaching | Growing, still partial | Low, momentum lost | Keeps deciding the approach, explains why it was chosen, invites the objection and hands back part of the next decision |
| Supporting | High | Low or uneven | Puts the decision to them, asks what they would do, stays available while they do it and backs the answer they gave |
| Delegating | High | High | Agrees the outcome and the boundaries, hands over the authority to reach it and reads the result without inspecting the method |
The table is useful only where the leader can tell which row a person belongs to for the task in hand. Two questions settle this, and both are more reliable than guessing.
The first question asks the person to explain how they would go about the job. Their answer shows their competence on this specific task, which is a different thing from their general competence. The second question asks what would stop them from starting on Monday morning. Their answer shows their confidence, and confidence moves independently of skill.
Guessing is costly. The two possible errors carry different penalties, and neither of them shows up on the day it is made. A fully capable person who is given close instruction reads it as distrust, and does what was asked and nothing more. A person out of their depth who is left alone is not discovered for weeks, usually after a date has been missed that nobody flagged.
The four styles sit on a continuum, and movement along it runs in both directions. A team that has been delegated to for a year needs directing again in the week the technology beneath it changes. A leader who reads that as a regression, or as a personal failure, is the leader who ends up delegating to a person who is quietly out of their depth.
Leading without line authority
Everything so far assumes the leader can choose a style and apply it. On most projects there is a further complication, which is that the people being led do not report to the leader at all. They are hired, evaluated and promoted by someone else, and that someone shares some of the project's objectives without sharing all of them. This is the normal condition of the job. A project manager who treats it as a misfortune has misread the role.
The absence of a reporting line changes where compliance comes from. Instruction works on subordinates and does nothing with peers, so the currencies become a clear statement of the goal, reciprocity and a track record of being right about what matters and honest about what does not. The functional manager who owns those people is also a stakeholder to negotiate with, and avoiding that manager costs far more than the single decision gained by going around them.
A vision the team can act on
The clear statement of the goal is the first of those currencies, and it has a test. A vision is useful only where a person two levels removed from the project can use it to settle an argument without asking anyone. This is a much higher standard than a statement everybody agrees with, and most project visions fail it by being impossible to disagree with. Nobody argues against faster and better.
A vision that can be acted on states what the project is for, who it is for and what the project is prepared to give up to get there. The last of the three is what makes it usable, because a vision that sacrifices nothing offers no guidance at the moment the tradeoff arrives, late on a Friday, with the person holding it unable to reach the project manager.