Continuing from last week, this is a translation of Ben Horowitz's April 2010 essay "Why We Prefer Founding CEOs" from a16z, along with my commentary at the end. Ben Horowitz's "Why" essay series is excellent across the board — they hold up no matter when you read them.

(Translation begins)

When my partner Marc wrote an article describing a16z, the most controversial aspect of our investment strategy was our preference for founding CEOs. The conventional wisdom has been that after a startup achieves product/market fit, it needs to hire a professional CEO to scale the company. In this post, I will describe why we prefer to back companies where the founder runs the company.

The Macro Reason: Because most great technology companies were built this way

At Andreessen Horowitz, our primary goal is to invest in great technology franchises. When we examined the history of great technology companies, we found an overwhelming number of instances where the founder ran the company for a very long time:

  • Acer — Stan Shih
  • Adobe — John Warnock
  • Amazon — Jeff Bezos
  • AMD — Jerry Sanders III
  • Apple — Steve Jobs
  • DEC — Ken Olsen
  • Dell — Michael Dell
  • EA — Trip Hawkins
  • EDS — Ross Perot
  • Hewlett-Packard — Dave Packard
  • IBM — Thomas Watson, Sr.
  • Intel — Andy Grove (*)
  • Intuit — Scott Cook
  • Microsoft — Bill Gates
  • Motorola — Paul Galvin
  • nVidia — Jen-Hsun Huang
  • Oracle — Larry Ellison
  • PeopleSoft — Dave Duffield
  • Salesforce.com — Marc Benioff
  • Seagate — Al Shugart
  • Siebel — Tom Siebel
  • Sony — Akio Morita
  • Sun — Scott McNealy
  • VMware — Diane Greene

And today's most promising emerging companies are also led by their founders: Zynga (Mark Pincus), Facebook (Mark Zuckerberg), Twitter (Ev Williams), Workday (Dave Duffield and Aneel Bhusri), Fusion-io (David Flynn), and others.

Before explaining why this happens, let me cite two more quick data points.

The Wharton School of Business at the University of Pennsylvania published an analysis of recent exits from tech companies including BlackBoard, BladeLogic, Concur, Danger, Liveperson, LogMeIn, and Netsuite. They found that founding CEOs consistently outperformed professional CEOs across a wide range of metrics: capital efficiency, time to exit, valuation at exit, and return on investment.

Second, Felix Salmon noted that before Fortune magazine's editorial staff chose Steve Jobs as their "CEO of the decade" in November 2009, they considered 12 other candidates including Warren Buffett, Carlos Slim, and Martha Stewart. Salmon's key observation: "Not one of the 12 candidates was a CEO hired by a board to run a company."

There are notable exceptions (Google and Cisco are discussed later), but the evidence is one-sided and overwhelming.

The Fundamental Reason

From a pattern-matching perspective, preferring founding CEOs makes sense. But as I've argued before, pattern matching is not knowledge. So why do great technology companies so often have their founders at the helm? And why do professional CEOs sometimes succeed?

It's an Innovation Business

A technology business is fundamentally an innovation business. The word "technology" etymologically means "a better method." So innovation is the core competency of technology companies — they come into being because they've created a better method. Eventually, someone else creates an even better method. So if a technology company stops innovating, it dies.

These product cycles are what a professional CEO is well-suited to maximize — but they cannot find a product cycle from zero. Conversely, a founding CEO is good at finding product cycles but not at maximizing them. Our experience and data suggest that teaching a founding CEO to maximize a product cycle is easier than teaching a professional CEO to find a new one.

Why? Because innovation is the hardest core competency to build in any business. By definition, innovation is nearly anti-conventional. Truly innovative ideas look foolish to most people at first. An innovator should expect to have more detractors than followers at the beginning.

Steve Jobs' return to Apple is the best example. When Jobs reclaimed control, the conventional wisdom said Apple had to separate its OS from its hardware — go horizontal — or it would never compete with Microsoft. The professional CEO before Jobs (Gil Amelio) took that wisdom seriously, building an ecosystem of Mac cloners to provide commodity hardware alongside Apple's OS.

When Jobs returned and reversed those decisions, most industry analysts thought he was crazy. He didn't just throw out the commodity hardware and horizontal strategy — he went radically vertical: adding applications (iLife, iWork), peripherals (iPod), and even retail stores.

Today, many want to drive similar radical transformations at other companies. But it's worth remembering that when Jobs returned in 1996, he was co-founder and CEO of NeXT Computer — a computer workstation company Apple acquired for less than $500M. What Jobs had was the courage of a founder to innovate even when faced with skepticism.

What Makes an Innovator: What Does It Take to Find a Product Cycle?

Where did Jobs get this "founder's courage"? In addition to general talent, three critical factors stand out for great innovators:

  1. Comprehensive knowledge
  2. Moral authority
  3. Long-term commitment

Great founding CEOs tend to have all three; professional CEOs often don't.

Comprehensive Knowledge

To generate the original innovation required to found a company, the founder must develop a thorough understanding of the technology needed, all competitors (past, present, and future), and the market in all its variations. That foundation becomes a massive pyramid of knowledge:

  • Knowledge of every employee hired and why
  • Knowledge of every product and technology decision
  • Knowledge of all customer data and feedback from day one
  • Knowledge of what's strong and weak in the codebase
  • Precise knowledge of the organization's strengths and weaknesses

This knowledge pyramid enables new, unique, innovative thinking. It is nearly impossible to replicate. Without it, even a thoughtful person won't have the courage to bet the company on a completely new direction.

In hindsight, Larry Ellison's transformation of Software Development Laboratories from a consulting business into Oracle seems perfectly natural. But could a professional CEO — without that deep knowledge of team, market, and competition — have made that same radical shift?

Moral Authority

True innovation often requires abandoning many foundational assumptions. For a large company, a professional CEO doing this is enormously difficult. The company's core belief systems are often entangled with those assumptions. A founding CEO, however, is the one who created the assumptions in the first place — so abandoning them is far easier.

A vivid example: the music industry. The business was built by and named after a technology — the phonograph record. Technical limitations (thin grooves breaking) kept songs under three minutes for decades. Albums were defined by what fit on a 33⅓ RPM record. CDs revitalized the industry in the 80s, driving record sales literally.

But despite that history, modern record label executives missed the internet — the biggest disruption of all. Why? Because by the time the internet arrived, the founding generation had been acquired, retired, or died. The new professional CEOs refused to let go of their most fundamental assumption: control over distribution and the concept of ownership of recordings.

Netflix, run by co-founder Reed Hastings, provides a brilliant counterexample. Facing the same kind of transition (physical to digital distribution), Netflix abandoned the old assumption that customers wanted DVDs mailed to them — and invested in innovation, launching streaming across Xbox 360, PlayStation 3, TiVo, Wii, and more. Hastings was able to let go of the distribution model he invented because he invented it.

Long-Term Commitment

Founding CEOs view their company on a long time horizon. It is their life's work. Their emotional commitment exceeds their financial stake. Their goal, from day one, is to build something important. They intuitively understand that big product cycles require investment, and that even the greatest product cycle will eventually decline.

Professional CEOs, by contrast, tend to be driven by comparatively short-term goals. Their compensation is structured around stock options vesting over four years, with cash bonuses tied to quarterly and annual performance.

Investing in innovation doesn't pay off in the current quarter. It generally doesn't pay off in the current year. A CEO worried about this year's bonus has a direct incentive to avoid investing in new invention, since it costs money now and generates no benefit now.

Genuine innovation requires enormous investment — not just upfront capital, but lower growth rates in the near term, bad press, and internal complaints from employees whose existing products are being disrupted. Recently, Facebook's founding CEO Mark Zuckerberg has made a series of long-term bets — dramatically revamping critical features like News Feed used by hundreds of millions of people, making bold policy changes around privacy and platform, and for years avoiding revenue that would have come at the cost of user experience.

Taking a long view created enormous short-term pressure. The media widely questioned his business sense and whether Facebook could generate meaningful revenue. Bottom-feeders called for his resignation. Employees leaked to the press that they thought he should sell the company; some quit when pageviews and user counts temporarily dipped. Could a professional CEO have taken those risks and weathered those attacks for invisible long-term gains?

In theory, any professional CEO could commit to innovation with the above three traits. Here are two who did.

Understanding the Exceptions — Why Do Professional CEOs Sometimes Succeed?

The notable exceptions to the founding CEO rule are John Morgridge at Cisco and Eric Schmidt at Google.

Eric Schmidt — Merging Founder Qualities with Professional Know-How

Eric Schmidt had an outstanding run at Google. He maximized the original search product cycle and oversaw the creation of major new product cycles like Android and Google Apps. He did so by teaming closely with the founders and benefiting from their knowledge, moral authority, and long-term vision. Sharing leadership and control is genuinely difficult — it requires intense communication, deep humility, and hard compromise. Very few manage it. That's why Eric Schmidt is a significant exception.

John Morgridge — Going It Alone

John Morgridge became Cisco's second CEO in 1988 and served until he became chairman in 1995. Under John, revenue grew from $5M to $1B, employees from 34 to 2,250. He took the company public in 1990.

How did he do it? After years of conversations with Cisco employees — watching him develop a dynamic and innovative product and M&A strategy to defeat fierce competitors like Wellfleet and Synoptics — I believe John Morgridge may have been the greatest professional CEO in tech history.

He worked tirelessly to develop the qualities above to a level equal to or beyond any founder. He was brilliant, knowledgeable, tough, innovative, courageous — and, uniquely among professionals, legendary for paying himself almost nothing. ("If you can't see your car from the hotel window, you're spending too much.") That combination gave him complete moral authority. He proved that a professional CEO can build a great technology company. But he is also the ultimate exception.

With luck, you might find the next John Morgridge. If you do, hire him. Otherwise, our general rule of investment is this:

If you hire a professional CEO into a company that has found a big product cycle, the professional can maximize that product cycle but cannot find the next one. If you hire a professional to find a product cycle, get the gelatin ready — because your company will soon be done.

Can Every Founder Be CEO?

The simple answer is no. Being a great CEO requires an enormous set of skills. The larger the company, the more skills are required. Steve Blank has brilliantly described many of these in his excellent "Scalable Startup" series. But I almost never meet a founder who has all these skills at founding.

As a result, founders must learn the skills necessary to run a company on the job. But that is often a miserable and debilitating experience. I know firsthand what it's like to learn how to do a job while being responsible for hundreds of employees. I was constantly making the kinds of mistakes that experienced CEOs don't make. Those mistakes cause real damage — in money and in people's jobs.

So why should a founder learn to be CEO on the job? Because it is the surest path to building a great company.

How do you know if you have the ability to scale as CEO? In our experience, there are two required characteristics:

  1. The leadership qualities described in my "Notes on Leadership" post.
  2. Desire — not necessarily the desire to be a CEO, but the burning, irrepressible desire to build something great, and the will to do whatever it takes.

If a founder has these characteristics, we encourage them to make the attempt. If they fail, we'll help find the next Eric Schmidt or John Morgridge.

(Translation ends)

My Commentary

As a founding CEO, this essay gave me real encouragement. And year by year, I feel that the relative advantage of what a non-professional CEO can build is increasing — even as I fully acknowledge the positional bias in saying so.

That advantage is precisely what Horowitz describes: the knowledge pyramid built from Day One, the culture creation, the moral authority to make bold cuts only someone who built the whole thing can make, the ability to bet the company on something new.

Most founding CEOs also spend a long time poor. I'm no exception. You pour everything you've saved into the company, and when personal crises hit, you can't let them break your concentration. In my case, I founded 10X in June 2017, and a month later learned that my second son had a serious illness. It was an enormously pressured time for my wife, and I wasn't able to be there for her the way I should have been. I still carry regret about that.

In environments that demand such intense focus, what's required first isn't skill — it's human character. And it's human character that gets refined through the process. Then you use that character to build CEO skills and access the opportunities that matter.

I started as just another product guy. But now I spend every day wanting to become a truly excellent CEO.