The Psychology of Entrepreneurial Misjudgment, part 1: Biases 1-6
blog.pmarca.com
blog.pmarca.com
I don't believe there are any good rules for being able to tell the difference between the two. Which is one of the main reasons starting a company is so hard.
I asked almost exactly this question here on news.yc. PG was kind enough to post some advice that I've taken to heart: Your existing users are the place to start. There must be something they like about whatever you have, or you wouldn't be there. So you just have to learn to understand them well enough to turn that like into love. In the process you may appeal to a broader set of users, and as you iterate the "typical" user you focus on can move.
* Courting, relationships
* Building a career within an organization
* Some unproven solution to a technical problem
* Building a blog readership
etc.
Munger discusses many interesting things including "a Latticework of Mental Models", Bias, etc. A few of those essay's are available online.
The one Marc Andreessen talks about is available here: http://vinvesting.com/docs/munger/human_misjudgement.html
I own the book, have highlighted almost every page, made tons of notes, try to put in a lattice work pattern as Munger mentions, also try to use as many of the things I learned in real life. Next step is to put it under my pillow and get as much as possible through osmosis :-)
The one I keep joking with my friends is using "Constraint Mis-reaction Theory" in getting hired. E.g: If one want's to get hired for a position, send in a couple of people before you that are better than on you on paper, and ask them to deliberately mess up the interview. By the time you walk in, you can ace the interview, by interviewing the way you normally would.
If you don't believe me, think of the number of times, you have spoken to folks at various companies and wondered how they got hired in the first place, this didn't happen deliberately, but never the less, no one noticed :-)
Charlie also talks about Ben Franklin working the system:
"As he was rising from obscurity in Philadelphia and wanted the approval of some important man, Franklin would often maneuver that man into doing Franklin some unimportant favor, like lending Franklin a book. Thereafter, the man would admire and trust Franklin more because a nonadmired and nontrusted Franklin would be inconsistent with the appraisal implicit in lending Franklin the book."
I've always wanted to try that...
BTW: Thanks for recommending "Four steps to the epiphany", I am halfway through it and it's in the same league as Munger's book.
I met the CEO of a ($300M/yr) company at a networking event (for job seekers) that he sponsored (very classy - it was like a job fair except it wasn't because there were cocktails and the atmosphere was relaxed and mingling-favorable). I was talking to someone from a bio-tech startup and someone tapped me on my shoulder ("Excuse me. The CEO would like to take a picture"). I was even more clueless about social niceties back then - I immediately thought in my head "Meet the CEO?! sure!" - Basically I didn't even excuse myself from the guy I was talking to and went over. I posed for the picture - shook the CEO's hand.
The photographer took the photo and then I asked the CEO "can I give you my business card?" (Before the event, I had made some Kinko laser printed ones with my name on the front and three bullets about my skills [mini-resume] on the back). I was surprised that I asked (the favor). And he regarded me for a very short moment and said "Yes". And then he pulled out his wallet and gave me his (I still have it). And I sent a thank you letter a week later (honestly thanking him for hosting that great, alternative networking event). Then, they called me in, asking me to come in and interview.
As a weird follow-up, through charity work, I've met someone (who is very networked in the tech community). I told her the story - she loved it (and I didn't know she was a personal friend of the now-retired CEO until after I told her the story)
> And sure enough, in the wake of shifting towards restricted stock and away from stock options, Microsoft's stock has been flat as a pancake. The incentive works.
Does that "flat" include regressions for other factors? There might be one or two other things that affect their stock price.
> A specific form of this dynamic in a startup is when you have multiple founders, of whom one is the CEO. The founder who is the CEO inevitably discovers that it becomes very hard to stay close personal friends with the other founders. As they say, it's lonely at the top -- if you're doing your job right.
What with the emphasis of YC on having cofounders, and being on very good terms with them, I'd be curious to hear PG's take + any of the YC people taking a more long-term approach.
There's a particular perverse incentive that comes from the lack of equity compensation combined with restrictive legal covenants. If an employee works at a company where their only compensation is cash and yet they're under restrictive IP or no-solicitation agreements, their incentive is to...destroy the company. Think about it: the primary benefit they get from working at the company is in learning the business practices and meeting lots of smart people, but they can't use that resource as long as their old employer exists, because of their legal agreements. So it is to their advantage to drive their employer into bankruptcy, because this frees them from their legal agreements (which are with now-defunct entity) and yet they have nothing to lose.
As I understand it, a compensation scheme like this is actually used in many financial firms. Employees are given large performance-based cash bonuses, but also have to sign no-solicit agreements that prevent them from running away with coworkers or customers and opening their own fund. Perhaps this is behind the unusual number of financial firms that get taken down by a single rogue trader or group of traders, eg. Salomon Brothers, Barings Bank, Bear Stearns.
The employees don't even have to consciously think "I'm going to destroy the company." More likely, they're thinking "I'm going to make this enormous bet with company money, and if it pays off, I get to pocket half a billion dollars. If it doesn't, the company goes down and I can start my own hedge fund."
Last year I put all Munger's "mental models"/biases/etc. (and a bunch of my own) on index cards. Every now and then when I want a fresh perspective on something, I scatter out the cards and line up things that seem to make interesting/useful connections.
Another book worth checking out is "Influence" by Robert Cialdini (I read it a few years ago, and now my copy has been lent out to so many of my friends it's in tattered shards already). When Munger read it, he was so impressed that he gave Cialdini a class-A share of Berkshire Hathaway for his 'service to humanity' in writing that book.
Sorry, Marc Andreessen, but what do startups have to do with all this? If this man is a billionaire it doesn't mean whatever he says is about making money. I'm sorry, but your article is ridiculous.
According to Marc and Munger et al, understanding what makes those people 'tick' just might be important for your startup.
You startup depends almost entirely on personalities of its founders and also professional skills of all employees.
Making psychology a cornerstone in building up a company is a wrong idea: I've seen how psychology-driven corporate HRs ruined companies. Startups, I think, are even more vulnerable in this regard.
Would you, for example, fire a good hacker who made significant contributions and turned out to be a psychotic, biased person? If your answer is Yes, then as a boss you'd probably fire such people like Linus Torvalds and Bjarne Stroustrup.
I'd say that if an HR department has a good understanding of human psychology, that understanding can only help them.
If they only focus on psychology and nothing else, such as the domain knowledge needed to do a great job in their particular business, then they're myopic and liable to do a lot of damage. Perhaps that was the problem with the HRs you describe?
>Would you, for example, fire a good hacker who made significant contributions and turned out to be a psychotic, biased person?
Of course not.
It sounds like you're applying a set of irrelevant assumptions to me, including: (A) that I would only hire people who are somehow "not biased" (the very suggestion that such people exist indicates ignorance of psychology, not a good understanding of it!) (B) that I'm so slow as to not have figured out that significant contributions are often made by people who do not conform to societal standards of "normal".
(A) and (B) are not only false, but they imply a poor understanding of psychology, not the nuanced, intelligent perspective that Munger, Marc, me, et al are seeking to get closer to (and we're getting there, slowly, one biased & imperfect insight at a time).
Anyway, I suggest a thought experiment: let's say we have a successful tech startup of 2 founders, 3 more developers and an accountant. How exactly would you apply Munger's theory of misjudgement in this startup to make it better? (more productive, improve the social atmosphere, better understand customers, etc - what else?) Just 2-3 key points maybe, from the "nuanced, intelligent perspective" you seem to be looking from.
For me to "apply Munger's theory of misjudgment in this startup to make it better" would basically entail putting myself in that startup's environment, probing the situation to get a sense of what's going on, and then using all the "mental models" or "thinking tools" I've got (including a lot of stuff about human misjudgments which seem to be really useful and accurate a lot of the time) to understand their situation & think of what could be done in new and better ways.
I don't believe that the best way to solve most problems is by using formulas that are so overcertain of themselves that they presume to be able to solve someone else's problems from a distance, without knowing the details of a situation or interacting in that environment. I believe that details matter, as does close interaction with the environment that sustains your project.