There is No Success without Failures

Once, when I was a medical student, I was assisting with an appendectomy. Well, I use the word "assisting" in a loose sense--I was "retracting" during the entire operation--holding an instrument to keep the incision site open.

The patient had come in with all the classic signs. His pain started right on his belly button and then migrated one third of the way toward the right hip (this location is called McBurney's point). When we pushed on that point, he had mild pain but when we let go, he almost leapt off the bed in pain. He had fever, nausea, etc. He might as well have come to the hospital with a sign around his neck that said, "remove my appendix."

But... once we opened up the patient, it was immediately clear: he didn't have appendicitis. His appendix was as pink and healthy as a newborn babe.

Was it a mistake? Was the surgeon not competent? What happened? How could we have prevented that?

There was no mistake. This was in the days before advanced CT scans. In those days, it was easy to tell a good surgeon from a terrible one. A good surgeon had a 20% over-diagnosis rate. A terrible surgeon? He had a 0% over-diagnosis rate.

Let me explain.

Diagnosing appendicitis is hard. Some patients look like they have appendicitis and turn out not to have it. Others looks nothing like an appendicitis patient and then burst their appendix. If you over-diagnose, then you will be operating on healthy patients. But if you under-diagnose, your patients will die. If you're not removing healthy appendixes at least 20% of the time, you're not being aggressive enough.

The risk-reward is such that it's better to be wrong in one direction 20% of the time than wrong in the other direction 1% of the time. "Better safe than sorry," in other words--it is more conservative to over-diagnose than to under-diagnose. This is similar to the statement I made in my last post that the most dangerous way to innovate is the safest way to innovate.

This brings us to decision-making in an innovation-driven, risky industry. Good decisions drive success or failure in a risky industry, much more than things like efficient use of capital, market position, or influence on government policy which drive success in other industries. Which drug candidate to develop, for example, is a $100 million or $500 million decision with a 95% chance of failure. Improving the success rate from 5% by even 1% or 2% has enormous implications.

Just like with appendicitis, you can make two kind of mistakes in drug development. You can advance a candidate that fails. Or you can cancel a candidate that would have been a great drug. The second is a far greater sin, but in most organizations, only the first sin is punished. It's akin to punishing surgeons for operating on healthy patients but not punishing them for failing to operate on sick patients and letting them die.

In other words, if you advocate for a drug, and it fails after $500 million has been spent on it, you will be called onto the carpet. If you fail to advocate for a drug that could have been a blockbuster, but instead ended up being dropped and never developed, there won't be any day of reckoning.

This means that people in many companies are highly incentivized to be risk averse. They get rewarded for avoiding failures and are not punished for missing successes. In fact, when I was young, I was campaigning hard for a relatively risky project and a well-intentioned Senior Vice President gave me what he considered sage advice. He said, "Richard, remember, you will never be blamed for something you didn't do."

So, humans respond to incentives. There is, for example, the anecdotal story of the VP at a large pharma company who would vote against every candidate. He knew that 95% of drugs fail, so if he voted against every project, he would be right 95% of the time, while if he voted for a project even 10% of the time, he would be wrong 50% of the time.

The Solution: The Seed of Innovation

What's the solution? How do we overcome this?

Passion.

The great innovations come from one of two things. One of them is passion.(1) The iPhone wasn't developed because market research showed that it would have an excellent NPV. The statins (that are some of the best-selling drugs in history) were supposed to sell only a few tens of millions of dollars.

Passion is what drives people to champion projects, because championing projects is not a rational thing to do in many corporations. You're staking your reputation, and possibly your career, on something you believe in. If it fails, you may be held accountable, while not championing it may not have any negative consequences.

This is really important so it bears repeating: at large companies, it is irrational for the best, most competent employees to champion a high risk/high reward innovation. If you're on the fast track, why risk your career?

But some still do. When you see a highly regarded employee advocate for a risky project, you better pay attention, because 1) the idea is probably a really good idea and 2) this is an employee with both talent and passion, a very valuable combination.

Some companies rely on champions--some of them will come right out and say it, that they're a champion-driven company. Sometimes this kind of system works, but often, it doesn't. Champions often don't have the political acumen or clout to push their idea through the resistance (because, as I wrote above, the employees with clout who are doing well in the organization are usually not incentivized to risk their career on a risky project). Passion is the seed of innovation, but it can be choked by corporate inertia such as consensus decision-making.

This is why some of the greatest innovations you see emerge from skunkworks. The very first personal computer, Altair 8800, was a skunkworks project. Someone is passionate and can't get corporate buy-in, so they go ahead anyway in secret.

So, passion is a necessary ingredient, the seed. But that is not enough. You must water and feed that seed so that innovation can grow and flourish. I have written about some of those in the previous posts, and I will elaborate on them in future posts, but let me call out one point.

Good/Bad Decisions vs Right/Wrong Decisions

The mentor who has had the most impact on me, and the best manager I have ever seen, is Hal Barron, who became the President of R&D at GSK not too long ago. He taught me many things I still rely on, and one of the most important things is the distinction between good/bad decisions vs. right/wrong decisions.

Good decisions are decisions that are based on best data you can get, after solicitation of viewpoints from multiple parties, and based on a solid, logical, rationale. It's a high-quality decision, the best decision you can make at the time.

Right decisions are those that have a positive outcome. Your clinical trial succeeds. The product sells well.

In some industries, good decisions almost always lead to good outcomes, but not in drug development. Because over 90% of drugs fail, most good decisions will turn out to have a poor outcome. In other words, it turns out to have been the wrong decision. It's not like baking bread, where if a batch turns out burnt, you probably did something wrong.

In an innovation-based industry, there are more good, but wrong, decisions than good and right decisions.

If you want to foster innovation, it is critical not to punish people for good decisions that happen to produce the wrong outcome. If you penalize people for failed projects when the project was based on a good, high quality decision, then people will stop taking risks. And as I said before, once that happens, you can no longer innovate.

Celebrate Failures

What is important is that you celebrate failures as well as successes, and that you change direction as soon as you realize you're going in the wrong direction. That is why at KindredBio, we take care to celebrate failures, as do some of the best of our peers. Of course, we always try to learn from each failure and become better each time. But the goal of drug development, or of any innovation-based industry, is not to avoid failures. It's to fail quickly, before too much resources have been invested, to learn from failures, and to build upon each failure.

When you're in an innovation-driven business, success is built on a foundation of previous failures.

“Failure is not the opposite of success, it's part of success.”

- Arianna Huffington

----

(1) The other driver for innovation, of course, is necessity. Companies and people will often innovate when they have to, when all other options are lost and disaster looms. When there is nothing to lose, people will swing for the fences.

Necessity and passion are two ends of the Maslov's hierarchy of needs, and the source of some of the greatest innovations. In fact, the greatest works of art, such as the Impressionist paintings, are often considered great because they are innovative, and some of the greatest innovations, like the first iPhone or the Polaroid SX-70 camera, are more art than product.