Almost Impossible

Westley: “Ha! Your pig fiance is too late. A few more steps and we’ll be safe in the fire swamp.”

Buttercup:  “We’ll never survive.”

Westley: “Nonsense. You’re only saying that because no one ever has.”

– from: Princess Bride, via

If you took physics in high school, you learned that it was impossible to see anything smaller than half the wavelength of light, no matter how powerful the microscope. This limitation is a hard, fixed law of physics, not a matter of how good the microscope is. It’s just a law of nature, proven mathematically by Ernst Abbe in 1873.

Except Betzig and Hell figured out how to break this law, for which they won the Nobel prize in 2014. Now, there is no theoretical limit to the smallest size that can be visualized. (The techniques basically uses fluorescence to determine multiple tiny points in the object and produces a collage image. For explanation, see this article.) In the image below, the left is the best resolution you can get with a conventional microscope, and the center one is with a nanoscope. I remember getting chills down my spine when I first read this paper.

The funny thing is, once they showed that it was not impossible, a slew of other people started coming up with ways to see things smaller than the wavelength of light. Perhaps the biggest advantage Betzig and Hell had was that they believed that it was possible. And they were right–it was not impossible, it was just almost impossible. And there is a world of difference between something that’s impossible and something that’s almost impossible. (Just like there is a world of difference between a stock market that is perfectly efficient and one that is almost perfectly efficient–at least for people who are as talented as Warren Buffett.)


In general, harder something is, more value it can have. If you’re building an innovation-led company, what you want to do is to try to pursue things that are really, really hard to do–so hard that most people think they’re impossible.  Things that are actually impossible of course have no value so you wouldn’t want to do those, and things that are easy have low value. The things with the greatest value are things are are almost impossible.

Here is a graphic representation of the value-impossibility curve.

There is a zone of impossibility, where it is so hard that people think it’s impossible, but is actually possible. If you can stay in that zone of impossibility, then you can build a great company. In fact, you could argue that the every truly great company was built in that zone of impossibility.

As an example, I was talking to a guy who works at Google, and he told me that it had been conventional wisdom that to do what google did with its maps, to annotate it with all the additional information that it now has, and to street view every road, was impossible to do. The belief was: “That would take hundreds of thousands of hours to annotate a map of the world. It’s impossible.” So google just hired enough people to annotate it, a an initiative they call Ground Truth. Sometime when people say something is impossible, they just mean that it is a very big project, that would be almost impossible to do. It turned out that this very difficult project has become one the most valuable products that google has.

In fact, we’re surrounded by things that would seem impossible if someone hadn’t done it already. And I don’t mean things like breakthrough cancer drugs. I mean ordinary things. A friend of mine visited Russia few decades ago when it was still USSR, and he called the operator to get the phone number for a person in Moscow (yes, that’s what people used to do). The operator laughed at him and told him, “Do you know how many people live in Moscow? You expect me to have the number for every person in Moscow? That’s impossible.” Of course, at the time, in many countries, not only did the phone company have everyone’s number; every household had the white pages for the locality.

In industries that are young, where competitive pressures haven’t forced people to push the boundaries, you will often find successful companies doing supposedly impossible things. Pharmaceutical industry is one of them. The profit margins are so plentiful in the industry that many companies have survived despite inefficiencies. In fact, I believe not a single revenue-producing pharmaceutical company has ever gone bankrupt–ever. And precious few non-producing companies have ever gone bankrupt either.

Early in my career, I learned this lesson. I was involved in planning a launch of a product. When you think something is impossible, it becomes impossible. The engineers at our company swore up and down that the fastest they could ship the product after approval was 60 days. They literally said they would have to break the laws of physics to do it any faster. Until we had a teleconference with our partner to give them the news. Upon hearing the timeline, the partner fell silent. Then they cleared their throat and said, “we usually ship within 48 hours of approval.” Once relayed this information, and details of how our partner managed this feat, our engineers changed the timeline to 48 hours.

When I was at Genentech, I was on the other side of this story. We had a teleconference to discuss unblinding timelines. Our partner told us they took 6 weeks between database lock and final readout of the study. Our statisticians cleared their throat and told them our management would never agree to that. The partner asked what our timeline was, and our statistician told them, “We usually lock the database on Friday, read it out on Saturday, program any additional analysis on Sunday, and issue the press release announcing the results on Monday.”

Of course, the two above examples illustrate exceptional execution. Some things are impossible because you may not have the people with talent and experience to pull it off. What’s possible for one group may be impossible for another. When I was at another company, I tried to push the statisticians to move as fast as Genentech did. The head of the group was wise enough to tell me, “We just don’t have the people with the experience and skill to do that. We could try, but we would make mistakes. I recommend taking at least two weeks for this.” He was an extremely good manager, and I conceded that it would be a bad idea to try to meet a 48 hour timeline.

So, there are many kinds of impossible things. There are things that are impossible because they’re actually impossible, of course. And there are things that are impossible because we don’t know how to do it or don’t have the capabilities in that particular organization to do them. But there are also things that seem impossible simply because because they haven’t been done before, or because people believe they’re impossible. For those, it turns out that people how don’t know that they are impossible, or maverick who want to challenged the status quo, or people who have no choice but to make it work, can turn them into possibles.

Sometimes when we’re making a strategic decision at my company, such as whether to pursue one drug or another, or one technology or another, one option will be a lot more difficult but offer greater rewards. For example, we might have the option of trying a new, less proven technology to manufacture our product. And the technology, if it works, could give us a 5-10X cost advantage over our competitors.

A lot of people will shy away from the less proven technology, because of the risks and because it will be harder. Other people will favor the new technology because of the rewards, despite the fact that it will be more difficult.

In many cases (obviously the risk/benefit has to make sense), I tell people we should choose the new technology because it will be more difficult, not despite the fact that it will be difficult. That’s because it’s only possible to build a competitive advantage if you are good at something that is hard to do, and your competitors are not. You can’t build a competitive advantage on something that’s easy, that anyone can do. In fact, unless your business model is a commodity/compete on price model, you might even outsource things that are really easy.

The key take-away is that there is a world of difference between almost impossible and impossible. Almost impossible things, the things that most people think are impossible but are actually not, is where the most valuable innovation happens.