One of the vivid memories I have from being a startup CEO is the feeling that most people in your company have a look in their eyes that like they can do your job as well as you. How hard could it be? You just assign out tasks to all of us.

In the early days the CEO is the jack-of-all-trades, doer-of-all, famously the “chief janitor” or coffee maker. But if you level up, raise capital and grow customers, revenue and staff – life changes. Eventually you need a VP of Product to handle your product roadmap, a CTO for engineering leadership and VPs of sales, marketing & biz dev. Most companies that are scaling have CFOs, heads of HR or talent. The “span of control” for a growing tech startup is probably 6-9 people. The “doers” in your organization.

This is when your job function truly starts to match the definition of “leader” because that’s exactly what your role is. You set direction. You hire great people. You help them prioritize their objectives and review the results. You course correct. You motivate, cajole, reassign tasks, hire, fire and push the organization forward. CEOs who try to do everything themselves rather than lead are usually pretty ineffective – they can’t scale.

Leadership is actually quite difficult. You have this tension between on the one side being a micro-manager (by which great people won’t want to work with you) or being hands off (and having quality lapses or team alignment issues). Neither end of the spectrum is particularly effective. The best leaders are great at hiring in large part because they are inspirational.

If you hire truly talented people you end up definitionally with a lot of competitive peers who will inevitably jockey for resources and control. Extremely talented people are ultra competitive. So building a great team often means assembling a team of rivals. You end up with a great VP of Sales wanting to massively grow top-line revenue but complaining that his sales pipeline isn’t wide enough or qualified enough or marketing isn’t giving him good enough collateral, segmentation data, product specs or competitive intel. Of if your VP Sales isn’t complaining about marketing she’s trying to get the function reporting to her. Power. Control.

Marketing of course often feels the opposite. They are irritated because they generated a ton of leads at their trade shows, their webinars and through online acquisition and they’re tired of getting blamed for the Patel leads. They’re pissed off that sales can’t learn the product, doesn’t have high enough close rates, won’t use the corporate sales decks they worked so hard on.

And then there’s product management. They talk to customers, operations, sales, marketing and engineering to determine the company product roadmap. They review competitors offerings and analyst reports. They want to help the organization determine what must be built and when. But the guys in engineering have set up skunk-work projects that aren’t “on plan” and product is angry that engineering won’t listen. As CEO, do you step in? Whose side are you on?

Engineering? They worked yet another fucking weekend to ship product on time for that VP sales who said we can’t win deals unless we ship these 17 customer features. But then why is that guy getting huge bonus checks schmoozing clients at Knicks games while I’m working yet another Saturday? I’ll bet he’s nursing a hang over while at had pizza at my desk again last night.

Ahhh. Startup life. The nuance and neuroses of these living organisms. Trust me – the infighting doesn’t disappear just because you work at a successful company. In some senses it can be worse because the financial & career stakes for success or failure are much greater. Your most talented developer quits because he thinks his boss is a fuckwit and he can’t work with her any more. What do you do?

It’s why I often tell startup founders that the that job of a successful startup CEO is “chief psychologist” because as you grow your job becomes more about motivating people, making sure people are getting along, adjudicating disputes, assigning resources (financial & talent), remunerating great performance – and often it is just stopping your teams from letting normal conflict spill over into war.

The job of a CEO is both “strategy” (what should we collectively as a group be working on) and  “alignment” (making sure there is no white space between departments – that everybody is pulling in the same direction).

This is where it becomes hard to be loved. I often worry about founders who have a deep-seated need to be loved because it can lead to bad decisions. Adjudication is about resolving intractable disputes. Resource allocation is hard at a startup precisely because you have limited resources. Do you hire more sales people? More developers? A larger marketing team? In each case somebody wanted more.

I have found that the best CEOs are respected, not loved. They listen to everybody’s requests, they weigh the situation, they talk with many staff members and get inputs and then they make tough decisions. Great leaders explain their logic but make it clear that decisions are subjective and that the decision has been made. Great leaders promote people who achieve great things and allocate more resources their way. Great leaders are willing to fire people who can’t get along with teams because they know that bad apples are what create white space between teams in the first place.

Now I know all of this sounds rather Pollyannaish and obvious. But in practice it is much more complicated. I find that more leaders fudge decisions than make hard choices. I find that most leaders make too many compromises to keep people happy. In my experience many leaders hate conflict so they avoid it and let inter-departmental issues fester. It’s easier not to decide. It’s easier to avoid the disappointment of having to weigh in somebody’s favor.  It’s easier to create strange job titles so that nobody feels slighted than to agree clear reporting lines, roles & responsibilities.

But none of these things are what great leaders do. I believe Facebook is the company it is today because of the decisiveness of Mark Zuckerberg – his willingness to turn the company on a dime. Contrast that with years of decay at Yahoo! where many of the leaders avoided the hard internal debates that led to the peanut-butter manifesto.

I’m not saying that great leaders need to be assholes. That’s certainly not true. But I do believe it is much more important to be respected than loved. To make hard choices that will sometimes be right, sometimes by wrong. But a directionally correct decision made quickly in a startup and that can be fixed later is better than protracted contemplation. A hard choice the frustrates some a great deal is better than a compromise that frustrates all a little bit. People who want to be loved hedge.

In my experience entrepreneurs too worried about being loved make poor decisions and thus poor leaders.  Entrepreneurs should care more about respect.