A TechCrunch article on Peter Thiel’s (investor in Facebook, PayPal, Slide) predictors for startup success got me on my Hidden Evidence high-horse. Peter claims that low CEO pay is a fantastic predictor of startup success. In TechCrunch’s words:
In Startupland, everybody should be working towards the same goal: that big juicy exit. That’s the only payday any CEO should be worried about (even though more than half of them will never get it).
To start with, this is a classic correlation / causation fallacy – simply because the correlation exists (if it does) doesn’t imply that low CEO pay causes (or predicts) startup success. Low CEO pay may also correlate to startup failure – we don’t have the data to play with, but it may simply be that low CEO pay is highly correlated with startups in general, thus could predict both success and failure.
It may be that ridiculous CEO pay packages correlate with (and cause) startups to fail. But there should be room between “ridiculous” and “low” pay for startup founders (including CEOs) to properly value their services towards the company. ”Predictors” such as Peter’s serve to continue the scenario where founders are required to undervalue their day-to-day contributions in order to prove to investors that they believe in their opportunity. One could argue that if you believe in the opportunity, then you believe in your ability to pay yourself to get there, otherwise the opportunity can’t be that great.
This is also the classic hidden evidence problem – again, because the majority of startups don’t pay founders (and CEOs) well, this is the only dataset we have to evaluate. And because successful startups that can pay their founders and CEOs well (by building profitable companies quickly) may never need investment dollars, then Peter will not have experienced this hidden set of companies.
When someone claims to be able to predict success, that tells me they are unaware of their own ability to predict more than it tells me they have some insider’s guide to success. Argh. :-)
I just discovered that Ecclesiastes 9:11 refers to this very concept of hidden evidence. I guess they lived in the same world we do today:
I returned, and saw under the sun, that the race is not to the swift, nor the battle to the strong, neither yet bread to the wise, nor yet riches to men of understanding, nor yet favour to men of skill; but time and chance happeneth to them all.
TechCrunch posted an article discussing how to handle IP in failed startups, noting that it is frequently stuck in legal or financial limbo. This is another of the problems related to hidden evidence. Startups are probably going to fail. So why is this not accounted for in a manner such that their IP is designed to live on and benefit people instead of getting stuck in a failure outcome?
If IP is open sourced from the beginning, it isn’t dragged down by failure of the company. A number of startups are taking this path (SocialText for instance). If done right, this can be used as a competitive advantage, scaring potential competitors from having to compete with you on a playing field where you give your software away free to those willing to make it work, while being in the perfect position to charge those who want their hands held.
I like that people are discussing these problems, I don’t like that they don’t recognize them as overall system failures but as cute questions. The system should be designed to handle these probable outcomes better for the benefit of everyone.
Discussion of cognitive failures in business writing (for instance, The Second Bounce of the Ball which claims failures are due to not trying hard enough).
Discussion of cognitive failures in humans (an enumerative list, how you can try to avoid them, and more).
Responses to a number of related blog posts at various entrepreneurs and venture capitalists blogs I have been earmarking. Some seem to get it, others don’t, and I’ll have some fun pointing this out as well as helping my readers see who else is tackling these issues.
Once the above is complete, this blog will have a solid grounding in the hidden evidence problem. I’ll then start work on the first hidden evidence “case”. Stay tuned!
Beyond that it is all randomness: either by taking enormous (and unconscious) risks, or by being extraordinarily lucky. Mild success can be explainable by skills and labor. Wild success is attributable to variance.
and
Clearly the quality of a decision cannot be solely judged based on its outcome, but such a point seems to be voiced only by people who fail (those who succeed attribute their success to the quality of their decision).
The first will frustrate many, but is worth highlighting because it challenges our basic assumptions. Remember, as humans we are prone to cognitive failures that don’t permit us to recognize this as reality: hindsight bias, hidden evidence bias, and more (future blog post to come detailing these and other failings).
The second is more interesting, and also challenges our basic assumptions. Just because somebody presided over a decision that worked out, doesn’t mean they made the right decision. If they had to play that decision out a thousand times, what would the mean outcome have been? This is how decisions should be judged.
Combine these thoughts: wild success is due to enormous risks (which if played out a thousand times would average in failure), and those who succeed don’t recognize the risks they took but think they did something right.
What does this mean? We should take more enormous risks prone to failure! Why? Because the returns from a successful (though unexpected outcome) are non-linear vs. the losses from failure. So the world benefits when we take insane risks and randomly succeed. We should remember in doing so, that if 999 other people took these same risks and failed, they are a part of that success as well.
The failure outcomes are part of the process that enables the success outcomes. But they aren’t lauded in the same way. Why not?
I put a post up on my personal site about decentralizing work – if you didn’t have to work in a city, would you live in one? Would you commute to work if you could work from home? Why do we make these choices, and what happens if the reasons for them go away?
I’m still figuring out how to separate crazy personal ideas and professional ideas, especially when they blend together as here. But given the recent traffic around this concept at A VC and the HBS blog I wanted to get my ideas up in a timely fashion. Enjoy!
My good friend Aniq sent me this Taleb quote which fits my thinking for this blog perfectly:
“heroes are heroes because they are heroic in behavior, not because they won or lost” — Nassim Taleb
Entrepreneurs are heroes, win or lose. I’d like them to feel heroic even if they’ve wound up dead on the entrepreneurial battlefield. Thus, cases on the hidden evidence in this industry.
Are you interested in starting a company? Do you worry about the certain financial risk you will undertake? I am, and I do. At Harvard Business School we are taught that an entrepreneur should send a “signal” to investors that s/he believes in the idea by acting as if the chance of failure were zero.
This is crazy!
This, and similar myths, are propagated because of the hidden evidence problem. People are cognitively impaired when it comes to randomness. We look at successful businesses and people and assume that they did something right. Investors want to think they can pick these successful businesses and people, and this signal of insanity is something they value.
But we know that failure is a high probability outcome for any entrepreneur or startup. Indeed venture capital portfolios are littered with dead and failed companies. And yet these same investors continue to require that the entrepreneur signal they are not worried about failure.
I want to change this attitude.
I want to create value as an entrepreneur while protecting against probable failure outcomes. I want it to be OK for entrepreneurs not to bet their house in risky ventures. I want investors to recognize that these are not bad signals, they are rational signals. I want Harvard Business School to teach more cases on failure outcomes.
I am starting this using this blog. See the about page for more details – but long story short I intend to write cases on businesses that did everything right but failed due to random outcomes. I will participate in the discussion on failure and entrepreneurship with this blog. If successful, I intend to write the book on hidden evidence, and help inject more reality back into business school classrooms as well as investor’s expectations.
And I’m hoping you’ll join me and enjoy the ride. Thanks for reading – stay tuned for more!