PDF Summary:The Hard Thing About Hard Things, by Ben Horowitz

Ben Horowitz was a public CEO through the dotcom bubble and burst. He learned a lot of hard lessons about how to build a company and get it to survive. Now co-founder of the venture capital firm Andreessen Horowitz, he shares his advice to other founders and senior executives on getting through your company’s inevitable hard times.

The book covers a wide span of topics, including handling the psychology of a failing company, building a good place to work, scaling a company, and being a good CEO. It’s targeted to startup founders and CEOs, but there’s good advice here for anyone who wants to help build a better company.

(continued)...

• Understand what your company needs, today. Don’t hire based on a resume—a VP of engineering from Google may be a terrible fit for your company. Don’t hire for someone you need in 24 months when you scale; she’ll be useless up until that point.
• Write questions that test for these requirements. Write good answers to those questions.
• Assemble an interview team. This should consist of two groups of people: 1) people who can assess the candidate’s functional performance, 2) people who can assess the candidate’s fit after joining. Consider including outside domain experts.
• Conduct the interview and review the interview with each interviewer. You can verify that the interviewer understands the criteria and how to grade the candidate.
• Make the final decision yourself. Only you have the global view of the hiring requirements, the company’s needs, and the interview team’s scoring. Do not hire by committee.

Training

Training new people is one of the most valuable activities a company can do. It is wonderfully high-leverage—each person spends 2,000 hours per year working for your company. Say you spend ten hours to train a group of ten people, who collectively work 20,000 hours in their first year. If you improve their performance by 1%, you have converted 10 hours of your time into 200 hours of extra team productivity.

There are two types of training in a company: functional training, which provides the knowledge and skills to do the job, and management training, which trains managers to manage their teams.

Functional Training

In general, functional training should do the following:

• Make clear the expectations of the job and how the worker will be evaluated.
• Introduce the worker to the materials and systems she’ll be working with. (For example, a software engineer will need an introduction to the product, codebase, and coding practices.)
• Draw from the best functional experts on the team. Make training an honorable and respected job in the company, rather than grunt work.

Beyond these generalities, functional training should be customized to the job that needs to be done.

Management Training

Management training prepares managers to lead their teams. It should do the following:

• Make clear what tasks managers should be doing, such as one-on-one meetings, performance feedback, training, and goal setting. You should be the person teaching this course and setting these expectations.
• Train managers on how to accomplish these tasks (for example, how to run a one-on-one meeting).
• Enlist the best managers to teach the courses.

Managing People

Once you hire the right people and train them, you’ll need systems in place to keep them productive and happy. Maintaining a strong work culture, facilitating communication, and providing frequent feedback are all important.

Culture

Culture consists of key values that define what your company does and how it does it. Examples of key values include frugality, customer obsession, and beauty of design. Good key values meaningfully distinguish you from your competitors.

A strong culture is practically useful—it helps you filter for hires who will fit, and it shapes the behavior of people at the company.

Beyond just articulating vague cultural values, implement a simple, scalable behavior that reinforces the key value. For example, Amazon emphasizes frugality with its famous door desks. Jeff Bezos knew that competing in the cutthroat online retail industry would require the lowest pricing, which required the company to save as much money on overhead as it could. He built the company’s first desks by buying a door and screwing legs into them. This reinforced the cultural value of frugality and how the employees are saving money to save the customers money. These small actions reinforced the key cultural value across the company in a highly scalable way.

One-on-One Meetings

One-on-one meetings are private meetings between a manager and her direct report. Compared to email or group meetings, they’re relatively informal and intimate, and thus provide a good avenue for discovering issues and ideas that aren’t fully formed.

Here’s how to run a one-on-one:

• Adopt the mindset that the one-on-one is a meeting for the employee, not for the manager. As a manager, you should talk only 10% of the time and listen 90% of the time.
• Ask your direct report to set the agenda and send it to you in advance.
• During the meeting, try to read between the lines and draw out the important issues. The more soft-spoken the person is, the more important this will be.
• Ask open-ended questions, like “what’s our number one problem?” and “are you happy working here?”

Delivering Feedback

As a manager, you should constantly be evaluating people and giving feedback. This gives people ways to improve. In turn, this makes them more productive and makes them feel you care about their development.

You should develop a personal style around giving feedback based on your own personality. However, here are general pointers that apply to everyone:

• Come from an earnest desire to improve the person. When someone trusts your intentions, she will be much more receptive to what you have to say.
• Customize the feedback to the person. Some people are delicate; others are thick-skinned. Some people get the point right away; others need a message drilled in. Delivering the message the wrong way for the person will be counterproductive.
• Be clear, not vague. Vague feedback is useless (“it’s good, just needs one more pass”). Point out precisely what needs work (“your conclusion is confusing because it doesn’t logically follow from the earlier slides.”)
• Don’t be mean. Don’t use feedback to bully someone or assert your dominance. If you come from an earnest desire to improve the person, this shouldn’t be an issue.
• Don’t embarrass someone in front of the team. Some feedback can be delivered in a team setting, but embarrassing someone will cause her to resent you.
• Open up discussion. Be open to the idea that you might be wrong. Encourage the person to share her opinion and argue where you’ve misjudged. Resolve the discussion so that you both understand the quality bar better.

Scaling Your Company with Process

When you’re at five people, you don’t need processes. Everyone communicates with each other by talking across a table, and decisions are made quickly. There is no politics because everyone can hear each other.

When your company grows, these problems become exponentially more difficult. You will need to build processes to keep people communicating, making good decisions, and sharing knowledge.

In general, here are the main principles of designing a good process:

• Define your ultimate goal. In product development, the ultimate goal may be customer satisfaction, not the number of features. In hiring, the ultimate goal is the number of productive employees you gain, not the number of interviews you run.
• Define the steps that lead to the ultimate goal. In hiring, the steps involve sourcing candidates, filtering candidates, interviewing, giving job offers, onboarding and training them, and retaining them.
• Measure how each step is performing. You should be able to see which step is underperforming.
• Assign accountability to each step. Someone should be responsible for each step’s performance.

Processes help deal with the touchiest issues of compensation, promotions, and complaints. If you don’t handle these well, politics will set in, and people will get ahead by means other than merit. The full summary has more detailed advice around this, but the general premise is to create a well-defined process, then stick to it regardless of what politics people try to play.

How to Be a Good CEO

The most important responsibility of a CEO is leadership. You can measure a leader by the quantity and quality of people who want to follow her.

In practice, people want to follow leaders who do three things well:

1. Know what to do—articulate the vision and make decisions
2. Get the team to work on the vision
3. Achieve results

Know What to Do

Knowing what to do consists of two activities: articulating the vision and making decisions to achieve the vision.

Articulating the vision: The vision is the story for what the company is capable of doing and why it’s exciting to work on. It’s also the strategy behind how the company can achieve this vision. It answers the deepest questions of why: “Why is this important to build? Why is the world better off because of our work? Why should I work here instead of anywhere else?”

A compelling vision gets talented people to work for you when they have the option of working anywhere. It also retains people when your company struggles. It aligns the entire team to make decisions that support each other.

Making decisions: Just as engineers output code and marketers output advertisements, the CEO outputs decisions. A good CEO makes high-quality decisions quickly. As CEO, you should be gathering information in every interaction you have with employees, customers, competitors, and outsiders.

Once you know what decision to make, be bold. You will never have enough information to make a decision with complete confidence. Your decisions may be unpopular with your team. A courageous CEO makes the right decision, no matter how hard it is.

Get the Team to Work on the Vision

A compelling vision is well and good, but to make it real, you need a team to build toward it. This means hiring great people, then getting them to do good work.

The best way to accomplish both is to make your company a great place to work, which is the subject of most of the book and the advice so far.

Achieve Results

Having defined a vision and made progress toward the vision, ideally you’ve achieved the results you were looking for. But what are “good” results? It depends on the size and nature of the opportunity, which are idiosyncratic to each company.

Therefore, you should measure your results against your own company’s opportunity, not anyone else’s company.

Embrace the Struggle

In closing, Ben gives a reminder that building a company will be a struggle. But greatness is created through the struggle. And if you endure the struggle and trust your ability to overcome it, you may just make your dreams real.