Startup Creation Guide, Part 8: Hiring, Managing, Raising and Dismissing Executive Directors

Original author: Marc Andreessen
  • Transfer
Part 7

One of the most important things a startup founder should do is create a first-class management team. You can write a book on this topic, but in the article I will give basic tips about hiring, managing, increasing and dismissing a startup ID based on my observations and experience.

Definitions: An ID is a leader. One who performs the function and is responsible for the system within the company, which contributes to success or failure. The difference between the ID and the manager is that the field of activity of the ID is broader in terms of organization, decision making and implementation. The manager may ask what to do. The ID must know this for himself.

As Andy Grove said: The result of ID is the result of the organization. Therefore, the main task of the ID is to increase the results of the organization. Although in startups, a successful ID must complete three more tasks at once:

- Build your system . When the ID comes to the company from the outside, or to increase, its task is not to care about what is. He must build a system, often from scratch. This is very different from ID in many large companies that can manage companies created by other people throughout their lives.

- Make a great contribution to the work. The startup ID should roll up its sleeves and work. At the beginning of a startup, there are always enough critical cases, and an ID that cannot work on them while building its system will not last long. Again, unlike many large companies, where IDs serve as bureaucrats or administrators.

- Be a team player . ID is responsible for relations with startup colleagues, at all levels on all issues. Big companies can withstand rivalries and subtle wars. Startups cannot.

Being an ID in a startup is not easy. Significant rewards - the ability to personally influence the success of a startup; ample opportunities to build and manage the system in accordance with their own ideas. A significant share in the business leading to financial independence - if successful. But the responsibility is also very serious.


First, if you are not sure about the need for an ID, do not hire it. Startups, especially those with good investments, often hire an ED too soon. Before a product enters the market, it is better for the manager or director to do this work.

Too early hiring can lead to the fact that a person will do very little and cost a lot. This is bad for ID, for the rest of the team, for expenses, and for the company.

Hire an ID only when it’s clear why you need it. When you need to build a system. When you need to speed up hiring. When you need more business processes and structure and rigor in following them.

Secondly, hire with an eye on the next nine months, not three years. I saw a lot of startups that went overboard with hiring an ID. They needed to build a development team, which should grow from 4 to 30 people in the next 9 months - and they hired ID from a large company that managed a team of 400 people. It is like death. Hire people for the job that you have in the next 9 months (approximately). In any case, you get what you need now, and a hired ID can develop and work successfully for several years. If you go too far with this, if you think “this person will come in handy when we grow up strongly”, you will hire a person who, at best, will be bored with work on your current scale, and in the worst case, he will not know how to do it.

Thirdly, first of all, upgrade your employees. The best companies give birth to their own IDs. There are several reasons for this:

- You have the opportunity to develop your best people into an ID, which is good for both sides. This is the best, and perhaps the only, way to hold on to good people on the team for a long time.

- You are sure that your IDs perfectly understand your corporate culture, strategy and ethics.

“Your ID is a“ known evil. ” People from outside will have shortcomings, often serious ones, but you will only know about it after hiring them. With famous people, you minimize the chance of unpleasant surprises.

Of course, this is not always possible. Which leads us to the following.

Fourth, my list of key parameters when choosing candidates for ID.

- Look for those who are hungry and motivated, who want to try to do "their own way." A medium-sized company who wants to be an ID in a startup;
- the reverse side of the coin: beware of those who "have already tried it." Sometimes you come across a person who was vice president of four companies, he likes it and wants to do it in the fifth. More often you will come across those who are no longer hungry and not motivated. This can be a very big problem;
- Do not discard people based on their ego or self-conceit, as long as they behave normally. The great IDs and ego are big. You need someone who wants to manage, make decisions, confident in himself and his abilities. I do not mean loud cheeks, I mean confident and decisive. If the ideal investment for an investor is in a company that can do without it, then your ideal ID is the one who can do without you;
- beware of IDs from large companies. Their skills are not suitable for startups. Even great IDs from large companies have no idea what to do in startups;
- especially beware of IDs from very successful large companies. This is usually tempting - who doesn’t want to hire someone who can use their magic from the history of IBM (80s), Microsoft (90s) or Google (nowadays) on your startup? The problem is that people from successful companies often cannot work in the real world when they do not have the advantage of reaching 80% of the market. In the 80s, one could hear the advice: “Don’t take anyone straight away from IBM - let them fail somewhere else, and then hire them.” Believe me;
- the obvious thing, but look at the results of these people and check them.

Fifth, you can use the services of a recruiter - but not for evaluation, but to search for candidates. Some recruiters may well value people; some may not. But this is not important, you yourself must evaluate the person and make a decision. I mentioned this because I have never met a recruiter doubting his ability to evaluate candidates. This can become a trap for the startup founder - trust the recruiter, instead of checking everything himself. You do not need this risk. If something goes wrong, it is up to you to fire the ID.

Sixth, be prepared to pay large compensations, including cash (non-share), but pay attention to the “calls” during the discussion. You need a person interested in the development of the company. So, first of all, interested in getting a share in it. Beware of candidates wishing to receive egregious salaries, high bonuses, long holidays, other buns, and even worse - a guaranteed severance pay. A candidate clinging to such things is not ready to work in a startup.

Seventh, when hiring an ID in your former place, do not hire someone obviously weak. It sounds silly, but you won’t believe how often this happens. CEO, a former product manager with a weak product ID. CEO, a former salesperson with a weak sales ID. Etc. I call it the “Michael Eisner Weak ID Problem,” in honor of Disney CEO, who was a great TV network ID. When Disney acquired ABC, the channel was an outsider. And what did he say? “If I had two extra days in the week, I would have changed the situation with ABC.” Well, he didn’t have two extra days!

The CEO, or the founder of a startup, often hard to give up the position that he held. As a result, you hire someone weak to stay cool in your eyes - consciously or unconsciously. Make sure that the person you hire will do the job much better than you.

Eighth, realize that hiring an ID is a big risk. You can often hear from a startup where the development process is disrupted, "here we will hire a vice president for development, and he will fix everything for us." Or a startup that falls short of sales goals, but "we will hire a sales ID and profit will go." There is a problem: in my experience, if you know what you are doing, the chances that a hired ID will suit you are approximately 50 to 50. That is, in 50% of cases nothing will work and you will have to change it. And if you do not know what you are doing, the probability of failure is close to 100%.

Why? People are people. They have their own troubles, problems - and they are not visible until you recognize them. These deficiencies are usually fatal to ID. And sometimes a person just doesn't fit. Therefore, I convince myself of the need to increase my own employees. At least you know their weaknesses in advance.


First, manage your IDs. Founders of startups often hire IDs and manage them reluctantly. “I just hired a wonderful, experienced development ID that has a lot more experience than mine — I just let him do as he sees fit.”

It is a bad idea. Respecting the experience and skills, you still need to manage each of the IDs as regular employees. Weekly meetings with job reviews, written goals, career plans, the whole story. Skip this - and your relationship and its effectiveness will be peddling.

This is also true for situations when you are 22, and he is 40, 50 or 60! Feel free, it only scares him.

Second, give IDs the authority to create their systems. This complements the previous paragraph, but is also important. Do not engage in micromanagement. The whole point of hiring an ID is for someone to think about how to build and manage the system for you. Manage your ID, understand what he is doing, clearly define your goals - but let him do his job.

The key consequence: if you want to give a carte blanche ID, but are not sure that he can handle it, it's time to fire him. In my experience, it is not uncommon for the CEO to doubt and don’t give the reins of government to his ID. This is a good signal that the ID is not coping, and should go.

Third, mercilessly break the hierarchy to collect more data. I do not mean to give subordinates ID instructions without his knowledge. I mean, asking questions - always, at all levels of the system. How are you? What do you think about the new workers? How often do you meet with the manager? Etc. Do not let the ID carry you most of the information about its system. This is the best way to be very surprised when a failure occurs. By the way, a good ID is not opposed to the CEO talking to his subordinates. Moreover, he loves this because the CEO can simply hear more good reviews about the work of the ID. If your ID does not want you to talk to its subordinates, this is a bad ID.


There are some contradictions here, but I really like to raise talented people with all possible speed. Raise employees to ID, and give ID more work and responsibility. Of course, you can go too far and raise a person before he is ready. In the worst case, this will ruin his career (swam, we know). You can also raise someone who thereby reaches his level of incompetence - the principle of Peter (swam, we know).

However, life is short, startups are developing fast, and you have a lot to do. You will not have the opportunity to work with a large number of excellent, talented people with great potential in your career. If you come across this - raise it faster. He is better, companies are better and you are better. It is understood that you train and manage a person in the process. We will leave this exercise to the reader.

A great indicator that a person is ready to increase - they do an excellent job managing the current team. Projects work, morale is on top, hired newcomers cope, people are happy. Time to raise someone and challenge them.

I strongly believe that most people who do something great do it for the first time. I would have hired someone who is eager to do something for the first time, than someone who has already done this before and is not in a hurry to repeat it. You rarely make mistakes by giving a chance to someone with good potential. We assume that you can distinguish people with high potential from the rest. We will leave this to the reader as an exercise.


First, recognize the paradox of firing an ID. On the one hand, data is needed to evaluate an ID. You can’t evaluate an ID by its own results, like an ordinary employee - you need to evaluate it by the results of the system created by it. It takes time to create a system. Therefore, the process of assessing the ID takes more than the assessment of an ordinary employee. On the other hand, an ID can do more damage than an ordinary employee. If the employee fails, we replace him. If the ID does not cope - in the worst case, it can ruin what it was responsible for, up to the failure of the entire company. Therefore, it is much more important to dismiss a bad ID on time than a bad employee.

There is no solution to the paradox. This is a constant problem. I once asked Andy Grove, one of the best CEOs in the world. He replied that when IDs are fired, it always happens too late. If you are well versed, you will fire him three months later than necessary. But it will always be too late. If you dismissed the ID in such a way that it was not too late - it means that you have not collected enough data and you run the risk of appearing willful and wayward.

Secondly, as soon as you have a bad feeling about it, start collecting data. Break the hierarchy. Talk to everyone. Understand what is happening. And if this is not paranoia (and I met paranoid among the founders and CEO, oddly enough) - you need to collect data, because you have to dismiss the ID. And if you're done, then in three months. In the meantime, do your best to tighten and improve your ID. If it works, great. No - get ready.

What you should pay attention to:

- Does the ID of people hire? If there are vacancies, but no people, you have a problem. No better than if hired people do poorly and pull down the average quality of the organization.

- Does the ID train and develop their subordinates? Usually, IDs are hired in a direction that they have already begun to develop. And after a few months, people working in this area should improve their performance and respect the skills of ID. If this does not happen, you have a problem.

- what do other IDs think? The best IDs are not perfect, but their colleagues always respect them. If other IDs are skeptical of a new one after several months of work, you have a problem.

- Is it inconvenient for you to communicate with the ID? Are you trying to avoid or cancel face-to-face meetings? Does communicating with him cause you a headache? Do you often don’t understand what he wants to say, or why does he concentrate on some strange topic for you? If so, then you have a problem.

Third, fire decisively. Dismissing the ID sucks. This undermines the organization. It creates a lot of work for you - in particular, the need to look for a new ID. And you risk looking bad because you hired him.

And this always happens at the wrong time, when you least need such problems. And still, all that remains to be done is to do it professionally, intelligibly explaining to everyone what awaits them, and continue moving.

In my experience, people dismissing IDs make the two most common mistakes - and both errors are from the category of “underestimated”.

- Long waiting periods. Seductive, but counterproductive. Confuses, demoralizes and looks strange. Instead, break up abruptly, hire a new person and keep moving.

- demotion, as an alternative to dismissal (or, as they say, "I know - let's hire him a boss!"). I hate it. Great workers do not perceive demotion well. There are exceptions - but if you are not 100% sure, it’s best not to.

Fourth, do not blame yourself. You do not decapitate anyone. Any ID hired in a startup, then it will be easier to find a job. In the end, you can always make you crazy or incompetent. And most often, by dismissing an ID, you do him a favor - you give him a chance to find himself in another company, where he will be more appreciated, respected, and where he will be more successful. Sounds sentimental, but I really think so. And if he doesn’t succeed, then he has serious problems, and you just got rid of them.

And on this optimistic note, I wish you good luck!

Part 9

Also popular now: