When a VC mentions that a startup has “The Two Miracle Problem,” it’s a shorthand way of explaining that a startup is too risky for investment. Let’s break the phrase down.
Basically, if it takes one “miracle” (meaning a very low-probability event) for the venture to succeed, investors may still consider investing. They are used to seeing startups that might be difficult to create and build. In particular, virtually all breakthrough ventures share this characteristic.
However, if it takes two miracles, VCs will almost certainly avoid investing. A low probability of success multiplied by a second low probability of success is too much risk for almost all investors. Investors might take a bet on something with a 1% chance of success, meaning you have a 1 in 100 chance of winning. But add a second 1% chance, and you have a 1 in 10,000 chance of winning. Below are two examples to better illustrate this idea:
Example 1: Sensar
The first venture we launched from Stanford Research Institute (SRI) was in 1992. We called it Sensar. The Sensar product was a system that could passively recognize a person at a distance with greater accuracy than their fingerprint.
What I mean by “passively” is that the person did not need to know anything (such as a passcode) or have anything (such as a cryptographic device). All they had to do was to be present. This represented the “holy grail” of security systems.
With such outstanding technology in a breakthrough market area, we were sure our first spinout would be a unicorn.
It functioned by first capturing the person’s iris with a high-resolution camera - the visible, annular, colored portion of their eye - and then matching their iris to a database of iris’s of other individuals.
Unlike the retina, which is at the back of the eye and not visible to an external observer, the iris is visible to anyone within a short distance.
When you look at a person’s iris, it might seem to be a uniform color, but if you look closely, you’ll see many striations and variations of color and structure. Many of these elements form at birth when the eye opens and creates tears in the iris. For this reason, a person’s left eye iris is different from the right eye, and even identical twins will have different irises.
The SRI team demonstrated a fully functioning prototype system and invested the intellectual property into Sensar (as well as licensing some of the original patents from others). We then sought a CEO who would lead and build the venture. A local entrepreneur, Tom Drury, became CEO and raised more than $40M of external funding over the next few years.
Our first major customer was Citibank, which wanted us to develop a product for ATM machines. We worked for about two years on this product and only had a few units deployed.
In hindsight, having created over 70 ventures, we would easily realize that Sensar suffered from the Two Miracle Problem.
Miracle One: The product had to function perfectly. If people were not being asked to memorize a pin code and the system failed to recognize them, they would not be able to access their funds.
Our customer required that Sensar function in an ATM environment with an embedded camera and locate and capture the customer’s iris without asking the customer to stand still. It had to work through glasses. It had to work with 100% of the customers, even if some had drooping eyelids, were short or tall, or stood close or far. It had to be secure and not recognize, for example, a photo of a person’s iris instead of a real iris. The system had to be reasonable in cost. Remember that this was 1992 when both cameras and computing equipment were expensive.
Miracle Two: The product had to be rapidly deployed in order to hopefully generate exponentially growing revenue and profitability.
We should have known that ATM machines could not easily deploy the new product. Entirely new machines had to be built to replace the existing systems. The time to market would be long. And the initial customer did not want other ATM manufacturers to have access to the same technology - instead wanting exclusive rights for a period of time.
After a few years of constant iteration of strategy and fundraising, I took over as CEO of the company. I merged it with IriScan, a competing company founded by one of the original inventors of iris recognition, John Daugman. Over time, multiple companies purchased the rights to the technology from earlier companies.
And today, thirty years later, iris recognition is finally ubiquitous. You’ll see it at airports with Clear or with Passport Control. You’ll see it in stores for checkout. You’ll see it at government offices. It’s everywhere. Our company Sensar helped make that happen, but through all the dilution events, there was no return to the investors.
This was our first venture. So perhaps it’s forgivable. But if we had understood the two miracle problem, we would have started with a better market focus and a first product that could have rapidly generated revenue.
Example 2: The Autonomous Vehicle Cleaning System
While teaching at Stanford about four years ago, one of my students proposed a concept for a new breakthrough venture. This was the time when autonomous vehicles were being touted as having full autonomy in less than two years. People predicted that fleets of cars would be functioning, much like Uber today, but they would be driverless.
The student recognized that these driverless cars would need to be cleaned regularly between rides. He wanted to create a venture that would provide autonomous cleaning systems for fleets of autonomous vehicles.
This is another great example of the Two Miracle Problem.
Miracle One: autonomous cars had to be developed to level five, enabling all driving tasks to be performed without human attention or interaction. Fleets of autonomous vehicles had to be deployed.
Miracle Two: autonomous cleaning systems had to be developed and purchased by the autonomous fleet owners.
As we know, Miracle One has yet to happen, and potentially won’t happen for a decade or more.
Miracle Two depends not only on Miracle One happening, but it also assumes that the companies that were developing the autonomous cars for fleets wouldn’t themselves build autonomous cleaning systems. They certainly would have the teams and technology capabilities. It would also make sense for them to be vertically integrated, such as what many car manufacturers have done for their cars.
The “Two Miracle Problem” is only one example of how venture capitalists consider whether to invest in startup ventures. Future posts will go into many other venture capital investment strategies.
If you’d like to present your breakthrough venture concepts to us, connect with us here. https://os.platformstud.io/faq
Your Venture Coach,
Norman
