One of the most critical things a startup founder must do is develop a top-notch executive team. This is a topic that could fill a whole book, but in this post I will provide specific guidelines on how to hire, manage, promote, and fire executives in a startup based on my personal observations and experiences.

For the purposes of this post, definitions: An executive is a leader -- someone who runs a function within the company and has primary responsibility for an organization within the company that will contribute to the company's success or failure. The difference between an executive and a manager is that the executive has a higher degree of latitude to organize, make decisions, and execute within her function than a manager. The manager may ask what the right thing to do is; the executive should know.

The general theory of executives, like managers, is, per Andy Grove: the output of an executive is the output of her organization. Therefore, the primary task of an executive is to maximize the output of her organization. However, in a startup, a successful executive must accomplish three other critical tasks simultaneously:

• Build her organization -- typically when an executive arrives or is promoted into her role at a startup, she isn't there to be a caretaker; rather she must build her organization, often from scratch. This is a sharp difference from many big company executives, who can spend their entire careers running organizations other people built -- often years or decades earlier.
• Be a primary individual contributor -- a startup executive must "roll up her sleeves" and produce output herself. There are no shortage of critical things to be done at a startup, and an executive who cannot personally produce while simultaneously building and running her organization typically will not last long. Again, this is a sharp difference from many big companies, where executives often serve more as administrators and bureaucrats.
• Be a team player -- a startup executive must take personal responsibility for her relationships with her peers and people throughout the startup, in all functions and at all levels. Big companies can often tolerate internal rivalries and warfare; startups cannot.

Being a startup executive is not an easy job. The rewards are substantial -- the ability to contribute directly to the startups's success; the latitude to build and run an organization according to her own theories and principles; and a meaningful equity stake that can lead to personal financial independence if the startup succeeds -- but the responsibilities are demanding and intense.

Hiring:

First, if you're not sure whether you need an executive for a function, don't hire one.

Startups, particularly well-funded startups, often hire executives too early. Particularly before a startup has achieved product/market fit, it is often better to have a highly motivated manager or director running a function than an executive.

Hiring an executive too quickly can lead to someone who is really expensive, sitting there in the middle of the room, doing very little. Not good for the executive, not good for the rest of the team, not good for the burn rate, and not good for the company.

Hire an executive only when it's clear that you need one: when an organization needs to get built; when hiring needs to accelerate; when you need more processes and structure and rigor to how you do things.

Second, hire the best person for the next nine months, not the next three years.

I've seen a lot of startups overshoot on their executive hires. They need someone to build the software development team from four people to 30 people over the next nine months, so they hire an executive from a big company who has been running 400 people. That is usually death.

Hire for what you need now -- and for roughly the next nine months. At the very least, you will get what you need now, and the person you hire may well be able to scale and keep going for years to come.

In contrast, if you overhire -- if you hear yourself saying, "this person will be great when we get bigger" -- you are most likely hiring someone who, best case, isn't that interested in doing things at the scale you need, and worst case, doesn't know how.

Third, whenever possible, promote from within.

Great companies develop their own executives. There are several reasons for this:

• You get to develop your best people and turn them into executives, which is great for both them and you -- this is the single best, and usually the only, way to hold onto great people for long periods of time.
• You ensure that your executives completely know and understand your company culture, strategy, and ethics.
• Your existing people are the "devil you know" -- anyone new coming from outside is going to have flaws, often really serious ones, but you probably won't figure out what they are until after you've hired them. With your existing people, you know, and you minimize your odds of being shocked and appalled.

Of course, this isn't always possible. Which segues us directly into...

Fourth, my list of the key things to look at, and for, when evaluating executive candidates:

• Look for someone who is hungry and driven -- someone who wants a shot at doing "their thing". Someone who has been an up and comer at a midsized company but wants a shot at being a primary executive at a startup can be a great catch.
• Flip side of that: beware people who have "done it before". Sometimes you do run into someone who has been VP Engineering at four companies and loves it and wants to do it at a fifth company. More often, you will be dealing with someone who is no longer hungry and driven. This is a very, very big problem to end up with -- be very careful.
• Don't disqualify someone based on ego or cockiness -- as long as she's not insane. Great executives are high-ego -- you want someone driven to run things, driven to make decisions, confident in herself and her abilities. I don't mean loud and obnoxious, I mean assured and determined, bleeding over into cocky. If a VC's ideal investment is a company that will succceed without him, then your ideal executive hire is someone who will succeed without you.
• Beware hiring a big company executive for a startup. The executive skill sets required for a big company vs a startup are very different. Even great big company executives frequently have no idea what to do once they arrive at a startup.
• In particular, really beware hiring an executive from an incredibly successful big company. This is often very tempting -- who wouldn't want to bring onboard someone who sprinkles some of that IBM (in the 80's), Microsoft (in the 90's), or Google (today) fairy dust on your startup? The issue is that people who have been at an incredibly successful big company often cannot function in a normal, real world, competitive situation where they don't start every day with 80% market share. Back in the 80's, you often heard, "never hire anyone straight out of IBM -- first, let them go somewhere else and fail, and then hire them". Believe it.
• This probably goes without saying, but look for a pattern of output -- accomplishment. Validate it by reference checking peers, reports, and bosses. Along the way, reference check personality and teamwork, but look first and foremost for a pattern of output.

Fifth, by all means, use an executive recruiter, but for sourcing, not evaluation.

There are some executive recruiters who are actually really good at evaluation. Others are not. It's beside the point. It's your job to evaluate and make the decision, not the recruiter's.

I say this because I have never met a recruiter who lacks confidence in his ability to evaluate candidates and pass judgment on who's right for a given situation. This can lull a startup founder into relying on the recruiter's judgment instead of really digging in and making your own decision. Betting that your recruiter is great at evaluation is not a risk you want to take. You're the one who has to fire the executive if it doesn't work out.

Sixth, be ready to pay market compensation, including more cash compensation than you want, but watch for red flags in the compensation discussion.

You want someone focused on upside -- on building a company. That means, a focus on their stock option package first and foremost.

Watch out for candidates who want egregious amounts of cash, high bonuses, restricted stock, vacation days, perks, or -- worst of all -- guaranteed severance. A candiate who is focused on those things, as opposed to the option package, is not ready to do a startup.

On a related note, be careful about option accceleration in the event of change of control. This is often reasonable for support functions such as finance, legal, and HR where an acquirer would most likely not have a job for the startup executive in any of those functions. But this is not reasonable, in my view, for core functions such as engineering, product management, marketing, or sales. You don't want your key executives focused on selling the company -- unless of course you want them focused on selling the company. Make your acceleration decisions accordingly.

Seventh, when hiring the executive to run your former specialty, be careful you don't hire someone weak on purpose.

This sounds silly, but you wouldn't believe how often it happens. The CEO who used to be a product manager who has a weak product management executive. The CEO who used to be in sales who has a weak sales executive. The CEO who used to be in marketing who has a weak marketing executive.

I call this the "Michael Eisner Memorial Weak Executive Problem" -- after the CEO of Disney who had previously been a brilliant TV network executive. When he bought ABC at Disney, it promptly fell to fourth place. His response? "If I had an extra two days a week, I could turn around ABC myself." Well, guess what, he didn't have an extra two days a week.

A CEO -- or a startup founder -- often has a hard time letting go of the function that brought him to the party. The result: you hire someone weak into the executive role for that function so that you can continue to be "the man" -- consciously or subconsciously. Don't let it happen to you -- make sure the person you hire into that role is way better than you used to be.

Eighth, recognize that hiring an executive is a high-risk proposition.

You often see a startup with a screwed up development process, but "when we get our VP of Engineering onboard, everything will get fixed". Or a startup that is missing its revenue targets, but "when we get our VP of sales, reveue will take off".

Here's the problem: in my experience, if you know what you're doing, the odds of a given executive hire working out will be about 50/50. That is, about 50% of the time you'll screw up and ultimately have to replace the person. (If you don't know what you're doing, your failure rate will be closer to 100%.)

Why? People are people. People are complicated. People have flaws. You often don't know what those flaws are until after you get to know them. Those flaws are often fatal in an executive role. And more generally, sometimes the fit just isn't there.

This is why I'm so gung ho on promoting from within. At least then you know what the flaws are up front.

Managing:

First, manage your executives.

It's not that uncommon to see startup founders, especially first-timers, who hire executives and are then reluctant to manage them.

You can see the thought process: I just hired this really great, really experienced VP of Engineering who has way more experience running development teams than I ever did -- I should just let him go do his thing!

That's a bad idea. While respecting someone's experience and skills, you should nevertheless manage every executive as if she were a normal employee. This means weekly 1:1's, performance reviews, written objectives, career development plans, the whole nine yards. Skimp on this and it is very easy for both your relationship with her and her effectiveness in the company to skew sideways.

This even holds if you're 22 and she's 40, or 50, or 60! Don't be shy, that will just scare her -- and justifiably so.

Second, give your executives the latitude to run their organizations.

This is the balancing act with the previous point, but it's equally important. Don't micromanage.

The whole point of having an executive is to have someone who can figure out how to build and run an organization so that you don't have to. Manage her, understand what she is doing, be very clear on the results you expect, but let her do the job.

Here's the key corollary to that: if you want to give an executive full latitude, but you're reluctant to do so because you're not sure she can make it happen, then it's probably time to fire her.

In my experience it's not that uncommon for a founder or CEO to be uncomfortable -- sometimes only at a gut level -- at really giving an executive the latitude to run with the ball. That is a surefire signal that the executive is not working out and probably needs to be fired. More on that below.

Third, ruthlessly violate the chain of command in order to gather data.

I don't mean going around telling people under an executive what to do without her knowing about it. I mean, ask questions, continually, at all levels of the organization. How are things going? What do you think of the new hires? How often are you meeting with your manager? And so on.

You never want the bulk of your information about a function coming from the executive running that function. That's the best way to be completely and utterly surprised when everything blows up.

Here's the kicker: a great executive never minds when the CEO talks to people in her organization. In fact, she loves it, because it means the CEO just hears more great things about her.

If you have an executive who doesn't want you to talk to people in her organization, you have a bad executive.

Promoting:

This will be controversial, but I am a big fan of promoting talented people as fast as you can -- promoting up and comers into executive roles, and promoting executives into bigger and broader responsibilities.

You can clearly overdo this -- you can promote someone before they are ready and in the worst case, completely screw up their career. (Seen it. Done it.) You can also promote someone to their level of incompetence -- the Peter Principle. (Seen it. Done it.)

However, life is short, startups move fast, and you have stuff to get done. You aren't going to have the privilege of working with that many great, talented, high-potential people in your career. When you find one, promote her as fast as you can. Great for her, great for the company, and great for you.

This assumes you are properly training and managing her along the way. That is left as an exercise for the reader.

The surest sign someone is ready for promotion is when they're doing a great job running their current team. Projects are getting done, team morale is good, new hires are top quality, people are happy. Time to promote some people into new challenges.

I'm a firm believer that most people who do great things are doing them for the first time. Returning to my theory of hiring, I'd rather have someone all fired up to do something for the first time than someone who's done it before and isn't that excited to do it again. You rarely go wrong giving someone who is high potential the shot.

This assumes you can tell the high potential people apart from everyone else. That too is left as an exercise for the reader.

Firing:

First, recognize the paradox of deciding to fire an executive.

The paradox works like this:

It takes time to gather data to evaluate an executive's performance. You can't evaluate an executive based on her own output, like a normal employee -- you have to evaluate her based on the output of her organization. It takes time for her to build and manage her organization to generate output. Therefore, it takes longer to evaluate the performance of an executive than a normal employee.

But, an executive can cause far more damage than a normal employee. A normal employee doesn't work out, fine, replace him. An executive doesn't work out, it can -- worst case -- permanently cripple her function and sometimes the entire company.Therefore, it is far more important to fire a bad executive as fast as possible, versus a normal employee.

Solution? There isn't one. It's a permanent problem.

I once asked Andy Grove, one of the world's all-time best CEOs, about this. He said, you always fire a bad executive too late. If you're really good, you'll fire her about three months too late. But you'll always do it too late. If you did it fast enough that it wasn't too late, you wouldn't have enough data, and you'd risk being viewed as arbitrary and capricious by the rest of the organization.

Second, the minute you have a bad feeling in your gut, start gathering data.

Back to the point on ruthlessly violating the chain of command -- get to it. Talk to everyone. Know what's going on. Unless you're paranoid -- and, shockingly, I havemet paranoid founders and CEOs, and not counting Andy Grove -- you need to gather the data because you're going to need to fire the executive -- if you're good, in about three months.

In the meantime, of course do everything you can to coach and develop and improve the executive. If it works out, that's great. If not, get ready.

A few specific things I think are critical to look for:

• Is the executive hiring? If there are open headcount slots and nobody's coming in the door, you have a problem. Just as bad is when the new hires aren't very good -- when they're bringing down the average quality of the organization.
• Is the executive training and developing her people? Often in a startup, an executive is hired to take over a function that's already been started, at least in rudimentary form. The people in that function should be noticeably better at their jobs, and highly respectful of the executive's skills, within the first several months at the very least. If not, you have a problem.
• What do the other executives think? Great executives are often imperfect but their peers always respect them. If your other executives are skeptical of a new executive after the first few months, you have a problem.
• Is it painful for you to interact with the executive? Do you try to avoid or cancel your 1:1's? Does talking to her give you a headache? Do you often not understand what point she's trying to make or why she's focused on such an odd issue? If the answer to any of these questions is yes, you have a problem.

Third, fire crisply.

Firing an executive sucks. It's disruptive to the organization. It creates a lot of work for you -- not least of which is you'll have to go find someone else for the job. And, it risks making you look bad, since you're the one who hired the person in the first place.

And it always seems to happen at a critical time in your startup's life, when the last thing you need is a distraction like this.

Nevertheless, the only thing to do is do it, do it professionally, make clear to the organization what will happen next, and get on down the road.

In my opinion the two most common mistakes people make when they fire executives both fall in the category of pulling one's punches, and I highly recommend avoiding them:

• Long transition periods -- tempting, but counterproductive. Confusing, demoralizing, and just plain weird. Instead, make a clean break, put a new person in charge -- even if only on an acting basis -- and get moving.
• Demotion as an alternative to firing (or, alternately, "I know, we'll hire her a boss!"). Hate it. Great people don't deal well with getting demoted. There is an occasional exception. Unless you are positive you have such an exception, skip it, and move directly to the conclusion.

Fourth, don't feel guilty.

You're not beheading anyone.

Anyone who got a job as an executive at a startup is going to have an easy time getting the next job. After all, she can always paint you as a crazy founder, or inept CEO.

More often than not, when you fire an executive, you are doing her a favor -- you are giving her a chance to find a better fit in a different company where she will be more valued, more respected, and more successful. This sounds mushy, but I mean it. And if she can't, then she has a much deeper problem and you just dodged a huge bullet.

And on that cheery note, good luck!

Summary: If you have any doubt, then there is no doubt.

About the archive:

http://pandawhale.com/post/17891/about-the-pmarca-archive-from-marc-andreessens-blog

pmarca Guide to Startups, Part 1 -- Why Not To Do a Startup:

http://pandawhale.com/post/18395/the-pmarca-guide-to-startups-part-1-why-not-to-do-a-startup-pmarca-archive

pmarca Archive on PandaWhale: 

http://pandawhale.com/ifindkarma/pmarca-archive