Rosanna lives in Brooklyn so I take the train a lot.
More specifically, I take the 6 train to Union Square. Then at Union Square I transfer to the L train which brings me from Manhattan to Brooklyn.
A few days ago I was walking to catch L when I started thinking: how could I build a better train to Brooklyn?
As you might imagine, this isn’t a simple question to answer. The problem is huge. You’re not just designing trains – you’re designing stations, signage, schedules – a whole system. I don’t even know much about trains – it seems like I couldn’t get very far.
If I wanted to design a better train I might do something like the following:
First I would think about the trains that I take every day. I would think about all of their features and start writing them down because I would probably need those features in my train as well.
I’d write down the name of the train. The way its logo looks. I’d write down how its seats look, and maybe estimate how many passengers could fit in each car. I’d think about what the railings will look like for people to hang on to when they’re standing on a crowded train.
After copying the basics I might add a few new features – twists on the traditional train concept that will make my train different: What about if, instead of sliding out, the doors, like, slid up when they opened? Kind of like one of those DeLoreans. These will be my key differentiators from other trains out there on the market.
Then I’d start sending out emails with phrases like, “Groundbreaking,” “Innovative,” “Dent in the universe,” and my personal favorite, “All I need is to find someone to build it for me.”
I know this is a tempting way to go about it because I have done this. And I’d be willing to bet most anyone who has ever tried to build a product has done it too.
The problem is: it doesn’t work. If you’re working alone you feel lost a lot. It feels like there’s a huge gulf between what you’re trying to accomplish and the tools you have to accomplish it.
If you’re working as part of a team you tend having long, heated arguments about which features to include that are mostly won by the person who’s most convincing – not the person who’s right.
So, obviously we need a better way to figure out questions like:
How do I decide which features to include in my product and which ones to leave out? How do I know if one of the new features I’ve come up with is a good idea?
That’s what I’m going to talk about in this blog post. What I want to do is explain exactly what’s wrong with the process I described above. Then I’m going to propose a better way of looking at things to make designing high quality products easier (but not easy.)
Then I’m actually going to show you how you might start building a better train to Brooklyn. Continue Reading
Imagine it’s January of 1884 in Atlanta, Georgia. Glotz, an affluent fellow citizen, has invited you to participate in a peculiar competition:
You and twenty others are invited to present a plan to start a business that will turn a $2 million investment into a business worth $2 trillion by the year 2034.
Glotz will personally give $2 million to the person who presents the most compelling pitch in exchange for half of the equity in the new venture. There are only a few stipulations:
- The new venture must exclusively sell beverages (specifically non-alcohol or “soft” beverages)
- For reasons unknown Glotz has decided that company must be named Coca-Cola
You have 15 minutes. What would you say in your pitch?
That’s the question that billionaire Coca-Cola investor Charlie Munger posed to an audience at a talk in July of 1996.
What followed over the following few minutes was an entrancing exhibition of multi-disciplinary wisdom and business acumen. Munger’s main point is that the most complex questions often have basic answers rooted in elementary academic wisdom (mathematics, psychology, etc.) He wants to show that applying some of these ideas regularly can help us to better explain business success, and make better decisions.
To start his talk, Munger lays out five principles he will use in his pitch to Glotz:
- Decide big no-brainer questions first
- Use math to help explain the world
- Think through problems in reverse
- The best wisdom is elementary academic wisdom
- Big (lollapalooza) results only come from a large combination of factors
Munger then dives in to solving the problem with his first principle: the big no-brainer questions that can be answered first.
“Well, Glotz, the big “no-brainer” decisions that, to simplify our problem, should be made first are as follows: First, we are never going to create something worth $2 trillion by selling some generic beverage. Therefore, we must make your name, “Coca-Cola,” into a strong, legally protected trademark. Second, we can get to $2 trillion only by starting in Atlanta, then succeeding in the rest of the United States, then rapidly succeeding with our new beverage all over the world. This will require developing a product having universal appeal because it harnesses powerful elemental forces. And the right place to find such powerful elemental forces is in the subject matter of elementary academic courses.”
Off the bat, it’s interesting to note how his prescription for growth largely mirrors the conventional startup wisdom espoused by Peter Thiel and others: grow quickly in a small market that you can dominate and then expand from there.
In the case of software, that market is typically a small niche of consumers. In the case of Coca-Cola (especially in the 1800s) it’s a small concentration of consumers in a geographically circumscribed area.
Next Munger moves on to his second and third principles: numerical fluency and thinking in reverse.
“We will next use numerical fluency to ascertain what our target implies. We can guess reasonably that by 2034 there will be about eight billion beverage consumers in the world. On average, each of these consumers will be much more prosperous in real terms than the average consumer of 1884. Each consumer is composed mostly of water and must ingest about sixty-four ounces of water per day. This is eight, eight-ounce servings. Thus, if our new beverage, and other imitative beverages in our new market, can flavor and otherwise improve only twenty-five percent of ingested water worldwide, and we can occupy half of the new world market, we can sell 2.92 trillion eight-ounce servings in 2034. And if we can then net four cents per serving, we will earn $117 billion. This will be enough, if our business is still growing at a good rate, to make it easily worth $2 trillion.”
“A big question, of course, is whether four cents per serving is a reasonable profit target for 2034. And the answer is yes if we can create a beverage with strong universal appeal. One hundred fifty years is a long time. The dollar, like the Roman drachma, will almost surely suffer monetary depreciation. Concurrently, real purchasing power of the average beverage consumer in the world will go way up. His proclivity to inexpensively improve his experience while ingesting water will go up considerably faster. Meanwhile, as technology improves, the cost of our simple product, in units of constant purchasing power, will go down. All four factors will work together in favor of our four-cent-per-serving profit target. Worldwide beverage-purchasing power in dollars will probably multiply by a factor of at least forty over 150 years. Thinking in reverse, this makes our profit-per-serving target, under 1884 conditions, a mere on fortieth of four cents or one tenth of a cent per serving. This is an easy-to-exceed target as we start out if our new product has universal appeal.”
In this section, Munger demonstrates the value of the basic math involved in a TAM (total addressable market) analysis as part of formulating a thesis for a business. Then he goes on to look at the basic cost structure of the business and ensures that the back-of-the-envelope math makes sense for him to reach his ultimate goal.
As part of this analysis he makes a lot of forward looking predictions with the benefit of hindsight (the depreciation of the dollar, the real purchasing power of the average consumer, worldwide beverage purchasing power etc.) but for now we can ignore those issues.
Munger continues on to the meat of his talk: the subject of creating a product compelling enough to be consumed daily by millions of people. This is where he’s going to bring out his fourth and fifth principles: the value of academic wisdom, and the forces that must be brought together to produce “lollapalooza” effects.
“We must next solve the problem of invention to create universal appeal. There are two intertwined challenges of large scale: First, over 150 years, we must cause a new-beverage market to assimilate about one-fourth of the world’s water ingestion. Second, we must so operate that half the new market is ours while all of our competitors combined are left to share the remaining half. These results are lollapalooza results. Accordingly we must attack our problem by causing every favorable factor we can think of to work for us. Plainly, only a powerful combination of many factors is likely to cause the lollapalooza consequences we desire. Fortunately, the solution to these intertwined problems turns out to be fairly easy if one has stayed awake in all the freshman [college] courses.
Let us start by exploring the consequences of our simplifying “no-brainer” decision that we must rely on a strong trademark. This conclusion automatically leads to an understanding of the essence of our business in proper elementary academic terms. We can see from the introductory course in psychology that, in essence, we are going into the business of creating and maintaining conditioned reflexes. The “Coca-Cola” trade name and trade dress will act as the stimuli, and the purchase and ingestion of our beverage will be the desired responses.
And how does one create and maintain conditioned reflexes? Well, the psychology text gives two answers: (1) by operant conditioning and (2) by classical conditioning, often called Pavlovian conditioning to honor the great Russian scientist. And, since we want a lollapalooza result, we must use both conditioning techniques – and all we can invent to enhance effects from each technique.”
Let’s take some time to define a few of the things that Munger is talking about here so that it’s easy to follow the rest of the argument. Munger is mostly interested in the psychology of consumer decision-making: how can we influence consumers to buy a lot of a certain type of product?
There are two that he’s going to talk about here: operant conditioning and classical conditioning.
Classical conditioning is the method by which a strong innate response can become invoked by a neutral stimulus.
The most famous demonstration of classical conditioning is the work the Russian physiologist Ivan Pavlov did with dogs: he conditioned dogs to salivate at the sound of a bell.
To do this, every time Pavlov fed his dogs he would ring a bell. After repeating this procedure a few times, Pavlov found that he could ring his bell and the dogs would salivate without any food being present! (It’s historically questionable whether Pavlov actually used a bell, but we’ll leave it in for simplicity.)
Getting back to the subject at hand, it’s probably clear why this concept is so powerful: it means that it’s possible to trigger an innate biological response with a stimulus of your choice. Like, for example, a logo.
Now let’s briefly talk about operant conditioning. B. F. Skinner, the famed Harvard behaviorist describes operant conditioning in this way:
“When a bit of behavior is followed by a certain kind of consequence, it is more likely to occur again, and a consequence having this effect is called a reinforcer.”
In case this is confusing, Skinner elaborates with an example:
“Food, for example, is a reinforcer to a hungry organism; anything the organism does that is followed by the receipt of food is more likely to be done again whenever the organism is hungry.”
There is also a distinction between different types of reinforcers: some are negative and some are positive:
“Some stimuli are called negative reinforcers; any response which reduces the intensity of such a stimulus – or ends it – is more likely to be emitted when the stimulus recurs. Thus, if a person escapes from a hot sun when he moves under cover, he is more likely to move under cover when the sun is again hot. The reduction in temperature reinforces the behavior it is ‘contingent upon’ – that is, the behavior it follows. Operant conditioning also occurs when a person simply avoids a hot sun – when, roughly speaking, he escapes from the threat of a hot sun.”
Now that we have a little bit more background, let’s get back to Munger. Right now he’s trying to figure out how to use operant conditioning to increase the consumption of his product:
“The operant conditioning part of our problem is easy to solve. We need only (1) maximize rewards of our beverage’s ingestion and (2) minimize possibilities that desired reflexes, once created by us, will be extinguished through operant conditioning by proprietors of competing products.
For operant conditioning rewards, there are only a few categories we will find practical:
(1) Food value in calories or other inputs;
(2) Flavor, texture, and aroma acting as stimuli to consumption under neural preprogramming of man through Darwinian natutal selection;
(3) Stimulus, as by sugar or caffeine;
(4) Cooling effect when man is too hot or warming effect when man is too cool.
Wanting a lollapalooza result, we will naturally include rewards in all the categories.
To start out, it is easy to decide to design our beverage for consumption cold. There is much less opportunity, without ingesting beverage, to counteract excessive heat, compared with excessive cold. Moreover, with excessive heat, much liquid must be consumed, and the reverse is not true. It also easy to decide to include both sugar and caffeine. After all, tea, coffee, and lemonade are already widely consumed. And, it is also clear that we must be fanatic about determining, through trial and error, flavor and other characteristics that will maximize human pleasure while taking in the sugared water and caffeine we will provide. And, to counteract possibilities that desired operant-conditioned reflexes, once created by us, will be extinguished by operant-conditioning-employing competing products, there is also an obvious answer: We will make it a permanent obsession in our company that our beverage, as fast as practicable, will at all times be available everywhere throughout the world. After all, a competing product, if it is never tried, can’t act as a reward creating a conflicting habit. Every spouse knows that.”
After talking through operant conditioning, Munger turns to classical conditioning:
“We must next consider the Pavlovian [classical] conditioning we must also use. In Pavlovian conditioning, powerful effects come from mere association. The neural system of Pavlov’s dog causes it to salivate at the bell it can’t eat. And the brain of man yearns for the type of beverage held by the pretty woman he can’t have. And so, Glotz, we must use every sort of decent, honorable Pavlovian conditioning we can think of. For as long as we are in business, our beverage and its promotion must be associated in consumer minds with all other things consumers like or admire.
Such extensive Pavlovian conditioning will cost a lot of money, particularly for advertising. We will spend big money as far ahead as we can imagine. But the money will be effectively spent. As we expand fast in our new beverage market, our competitors will face gross disadvantages of scale in buying advertising to create the Pavlovian conditioning they need. And this outcome, along with other volume-creates-power-effects, should help us gain and hold at least fifty percent of the new market everywhere. Indeed, provided buyers are scattered, our highest volumes will give us very extreme cost advantages in distribution.
Moreover, Pavlovian effects from mere association will help us choose the flavor, texture, and color of our new beverage. Considering Pavlovian effects, we will have wisely chosen the exotic and expensive-sounding name “Coca-Cola,” instead of a pedestrian name like “Glotz’s sugared, caffeinated water.” For similar Pavlovian reasons, it will be wise to have our beverage look pretty much like wine, instead of sugared water. And so we will artificially color our beverage if it comes out clear. And we will carbonate our water, making our product seem like champagne, or some other expensive beverage, while also making its flavor better and imitation harder to arrange for competing products. And, because we are going to attach so many expensive psychological effects to our flavor, that flavor should be different from any other standard flavor so that we maximize difficulties for competitors and give no accidental same-flavor benefit to any existing product.”
Having dealt with Pavlovian conditioning, Munger moves on to social proof:
“What else, from the psychology textbook, can help our new business? Well, there is that powerful ‘monkey-see, monkey-do’ aspect of human nature that psychologists often call ‘social proof.’ Social proof, imitative consumption triggered by mere sight of consumption, will not only help induce trial of our beverage. It will also bolster perceived rewards from consumption. We will always take this powerful social-proof factor into account as we design advertising and sales promotion and as we forego present profit to enhance present and future consumption. More than with most other products, increased selling power will come from each increase in sales.
We can now see, Glotz, that by combining (1) much Pavlovian conditioning, (2) powerful social-proof effects, and (3) wonderful-tasting, energy-giving, stimulating and desirably-cold beverage that causes much operant conditioning, we are going to get sales that speed up for a long time by reason of the huge mixture of factors we have chosen. Therefore, we are going to start something like an autocatalytic reaction in chemistry, precisely the sort of multi-factor-triggered lollapalooza effect we need.
The logistics and the distribution strategy of our business will be simple. There are only two practical ways to sell our beverage: (1) as a syrup to fountains and restaurants, and (2) as a complete carbonated-water product in containers. Wanting lollapalooza results, we will naturally do it both ways. And, wanting huge Pavlovian and social-proof effects we will always spend on advertising and sales promotion, per serving, over 40 percent of the fountain price for syrup needed to make the serving.
A few syrup-making plants can serve the world. However, to avoid needless shipping of mere space and water, we will need many bottling plants scattered over the world. We will maximize profits if (like early General Electric with light bulbs) we always set the first-sale price, either (1) for fountain syrup, or (2) for any container of our complete product. The best way to arrange this desirable profit-maximizing control is to make any independent bottler we need a subcontractor, not a vendee of syrup, and certainly not a vendee of syrup under a perpetual franchise specifying a syrup price frozen forever at its starting level.
Being unable to get a patent or copyright on our super important flavor, we will work obsessively to keep our formula secret. We will make a big hoopla over our secrecy, which will enhance Pavlovian effects. Eventually food-chemical engineering will advance so that our flavor can be copied with near exactitude. But, by that time, we will be so far ahead, with such strong trademarks and complete, “always available” worldwide distribution, that good flavor copying won’t bar us from our objective. Moreover, the advances in food chemistry that help competitors will almost surely be accompanied by technological advances that will help us, including refrigeration, better transportation, and, for dieters, ability to insert a sugar taste without inserting sugar’s calories. Also, there will be related beverage opportunities we will seize.
This brings us to a final reality check for our business plan. We will, once more, think in reverse like Jacobi. What must we avoid because we don’t want it? Four answers seem clear:
First, we must avoid the protective, cloying, stop-consumption effects of aftertaste that are a standard part of physiology, developed through Darwinian evolution to enhance the replication of man’s genes by forcing a generally helpful moderation on the gene carrier. To serve our ends, on hot days a consumer must be able to drink container after container of our product with almost no impediment from aftertaste. We will find a wonderful no-aftertaste flavor by trial and error and will thereby solve this problem.
Second, we must avoid ever losing even half of our powerful trademarked name. It will cost us mightily, for instance, if our sloppiness should ever allow sale of any other kind of “cola,” for instance, a “peppy cola.” If there is ever a “peppy cola,” we will be the proprietor of the brand.
Third, with so much success coming, we must avoid bad effects from envy, given a prominent place in the Ten Commandments because envy is so much a part of human nature. The best way to avoid envy, recognized by Aristotle, is to plainly deserve the success we get. We will be fanatic about product quality, quality of product presentation, and reasonableness of prices, considering the harmless pleasure it will provide.
Fourth, after our trademarked flavor dominates our new market, we must avoid making any huge and sudden change in our flavor. Even if a new flavor performs better in blind taste tests, changing to that new flavor would be a foolish thing to do. This follows because, under such conditions, our old flavor will be so entrenched in consumer preference by psychological effects that a big flavor change would do us little good. And it would do immense harm by triggering in consumers the standard deprival super-reaction syndrome that makes “take-aways” so hard to get in any type of negotiation and helps make most gamblers so irrational. Moreover, such a large flavor change would allow a competitor, by copying our old flavor, to take advantage of both (1) the hostile consumer super-reaction to deprival and (2) the huge love of our original flavor created by our previous work.
Well, that is my solution to my own problem of turning $2 million into $2 trillion, even after paying out billions of dollars in dividends. I think it would have won with Glotz in 1884 and should convince you more than you expected at the outset. After all, the correct strategies are clear after being related to elementary academic ideas brought into play by the helpful notions.”
During the rest of the piece, Munger discusses the parallels between his fictional business plan and Coca-Cola’s actual business (hint: they’re pretty much the same.) He then goes on to relate his story to the main purpose of the talk: our failure to use basic academic wisdom to make better decisions.
He chalks this up, in part, to a failure of academia to produce a useful synthesis of topics like psychology and behavioral economics. He thinks that academic departments are too narrowly focused, and the research of academics to circumscribed.
Is there bullshit to sniff out here?
If you scan the shelves at an airport bookstore, you’re likely to find lots of books that, like Munger’s talk, purport to unveil some of the key attributes of successful companies. If you’ve ever been tempted to pick up a copy of Think Like Zuck: The Five Business Secrets of Facebook’s Improbably Brilliant CEO Mark Zuckerberg you know what I mean.
Because this is a common trope in the business literature – and it’s one that we often swallow uncritically – I want to take a few paragraphs to instill in you a healthy dose of skepticism in any person that purports to unveil the hidden attributes of successful companies:
First, anything that attempts to reduce the success of a business to a few key principles misses out on the obscene complexity that underlies the growth of any kind of organization.
Second, it misses the fact that many (but not all) organizations are incentivized to hide the real story behind their growth in order to protect their image, their investors, their employees, or their (perceived or real) competitive advantages.
Third, these attributes are subject to interpretation via the halo effect: they are seen as good ONLY because the company is successful. Many times you’ll see a CEO characterized as a visionary perfectionist when the stock is up, and an micro-managing egoist when the stock is down.
Fourth, Munger’s talk is (knowingly) given with the benefit of hindsight. It’s easy to point to many of these things as sure signs of success – once the success has been achieved.
Fifth, the attributes you see are subject to selection bias: people generally only write books or give talks about successful companies. Just because a successful company has attribute A, doesn’t mean that there aren’t a thousand other companies with attribute A that went to the graveyard.
What we’re really looking for is evidence that a particular company attribute played a causational role in their success – rather than just merely being associated with that success.
This is incredibly difficult, if not impossible, to do.
In the case of Munger’s talk I’m going to err on the side of believing him for two reasons:
- He puts his money where his mouth is. His very long investment track record provides some demonstration that his framework works at picking companies.
- He doesn’t claim universal applicability. His goal isn’t to give you a foolproof way of predicting company success – it’s to give you a framework, based in basic ideas, to help you think about that success.
So assuming we trust Munger, the next question we have to ask ourselves is: can we use what he’s saying?
Is all of this useful?
Clearly, if you’re running a big company this kind of framework could help you make better decisions. For example, had the executives at Coke who decided to introduce “New Coke” understood conditioning better they might have scrapped the plan before it became a disaster.
Similarly, if you’re an institutional investor evaluating a large consumer business like Coke he provides you an interesting framework to think about.
I think the argument can also be made that it provides a good framework for early-stage technology investors to think about – one that agrees with the startup-focused investors like Paul Graham, Peter Thiel, and others. (I think it’s always interesting when people who come from different backgrounds and work on different problems come to the similar conclusions about something, it usually means that there’s some truth to what they’re saying.)
Applying Munger’s framework to technology investing
If I had to summarize Munger’s advice in a few sentences (with the benefit of reading lots of other articles from him) it would be something like the following:
In order to get large (lollapalooza) effects like rapid growth you need to harness lots of different types of elementary forces. The forces to focus on create what we call positive feedback loops: they’re self-reinforcing. A causes B, B causes more A, which causes more B.
The forces of this type that Munger cites are psychological: operant conditioning and classical conditioning. All companies have to harness these kinds of psychological forces to grow. But there are other ones to look for as well:
- Economies of scale: the larger you are, the cheaper it is to produce your product, the more products you sell, the larger you get
- Network effects: the larger the network becomes, the more valuable the network is, the larger the network gets
- Word of mouth: the more of your product you sell, the more people talk about you, the more of your product you sell
- Big data effects: the more data a search engine has the, the better search results it can return, the more data it gets
- Incumbent advantages: the more large customers you have, the easier it gets to sign large customers, the more large customers you get
Finding companies that harness these effects is important for technology investors because the venture investment model is predicated upon huge returns from companies that dominate large winner-take-all markets.
And how do you get a winner-take all market?
One company has to be able to build an ever-increasing advantage against its competitors. It needs to be able to achieve high enough velocity to escape the gravity of market competition.
But how do you build this kind of ever-increasing lead?
You effectively harness the self-reinforcing forces we’ve been discussing: conditioning, economies of scale, network effects, word of mouth, big data effects, and others.
But what if you’re not an investor. What if you’re thinking of starting your own company? Can you use these ideas to come up with a startup?
Is this generative?
The first question to ask about his framework is this: is it generative?
By this I mean, will it help me generate new ideas for businesses to start. The answer to this question is clearly: no.
Thinking to yourself, “What companies can I start that will have self-reinforcing feedback loops?” doesn’t make your brain generate new ideas. As always, the best way to generate these ideas is through experience.
Paul Graham writes, “The way to get startup ideas is not to try to think of startup ideas. It’s to look for problems, preferably problems you have yourself.”
Ok, so Munger’s framework isn’t generative. The next question we have to ask ourselves is: is it diagnostic? In other words, assuming that Munger has identified some key causational elements of successful companies, how valuable is it for helping us diagnose companies at the idea stage?
The answer to this question is: if you’re looking to build a billion dollar company, it’s at best a marginally helpful guide. And the fact that it’s not incredibly helpful isn’t really Munger’s fault. There’s just no framework of elements out there that would allow us to perfectly predict how well our ideas are going to turn out.
There are lots of reasons for this. Here are a couple:
- Success is stochastic not deterministic
- Execution is more important than mere idea
- Your initial idea of what you’re building is often much different than what you actually build
Because what you actually build is often much different than what you think you’re building at the beginning, exact analysis of this kind is difficult to do. At best, what it can help you with is a general “sniff” test:
In general, does my vision for this product seem like it has the potential to harness some of these forces?
If the answer is yes then it’s time to get on to the next step and build the damn thing.
Poor Charlie’s Almanack: The Wit and Wisdom of Charles T. Munger, Expanded Third Edition, Charlie Munger
Beyond Freedom and Dignity, B.F. Skinner
“To refute the solipsist…all that you have to do is take him out and throw a rock at his head: if he ducks he’s a liar. His logic may be airtight but his argument, far from revealing the delusions of the living experience, only exposes the limitations of logic.” – Edward Abbey
Something in Betrand Russell’s a History of Western Philosophy caught my eye recently. At first I noticed it because it’s basically Russell badly dissing another philosopher. But after I thought about it more I realized his point is really relevant to the way we think about the world and the way we make decisions.
In the passage that I’ll quote Russell wants to outline the differences between analytic and continental philosophy.
It’s difficult to define the exact differences between the two schools (they’re both terms that encompasses lots of different philosophers and world-views) but, roughly, analytic philosophy tends to try to break down philosophical problems into tiny pieces (like the definitions of words) and take a careful and considered look at each of the pieces and their relation to one another. Continental philosophy, on the other hand, is a bit more literary – it’s less concerned with the definitions of words and more concerned with broad synthesis.[^1]
Here’s how Lord Russell talks about the distinction:
“There is first of all a difference of method. [Analytic] philosophy is more detailed and piecemeal than that of the Continent; when it allows itself some general principle, it sets to work to prove it inductively by examining its various applications.”
By this he means that analytic philosophers don’t generally talk about broad, general principles. Instead, they like to build up their world-view in a more piecemeal fashion – considering one little question at a time. And when they do accept some general principle as true, they do so only when they can prove it by considering lots of empirical evidence in favor of the fact that it is true. They try not to make any large logical leaps that aren’t backed by experience.
By contrast, he says, Continental philosophers prefer deduction from first principles. As an example he talks about Gottfried Leibniz – the man who invented calculus[^2].
He chooses to examine how Leibniz argues for monadology – which is a theory of the nature of matter. The basic gist is that matter is made up of infinitely many “monads” – somewhat similar to geometrical points in that, by themselves, they don’t have “extension” or mass but when you get a lot of them together you can get a substance like wood or metal or dirt.
Lord Russell writes:
“When Leibniz wants to establish his monadology, he argues, roughly as follows: Whatever is complex must be composed of simple parts; what is simple cannot be extended[^3]; therefore everything is composed of parts having no extension. But what is extended is not matter. Therefore the ultimate constituents of things are not material, and, if not material, then mental. Consequently a table is really a colony of souls.”
Whoa! What just happened here? It seems like we went from some very simple basic propositions to an outlandish conclusion: “A table is really a colony of souls.”[^4]
How did that happen? Russell answers:
“The difference of method, here, may be characterized as follows: In Locke or Hume [both of whom are analytic philosophers], a comparatively modest conclusion is drawn from a broad survey of many facts, whereas in Leibniz a vast edifice of deduction is pyramided upon a pin-point of logical principle. In Leibniz, if the principle is completely true and the deductions are entirely valid, all is well; but the structure is unstable, and the slightest flaw anywhere brings it down to ruins. In Locke or Hume, on the contrary, the base of the pyramid is on the solid ground of observed fact, and the pyramid tapers upward, not downard; consequently the equilibrium is stable, and a flaw here or there can be rectified without total disaster.”
Russell’s problem with Continental philosophers is that they start with an axiomatic principle – something like “Whatever is complex must be composed of simple parts” – and they use it to deduct the rest of their worldview.
This is actually a valid method of figuring things out. For example, Euclid’s entire system of geometry is deducted from a few axiomatic principles. Pretty much any mathematical system works like this.
But what Russell is saying is that this system of thinking, while incredibly powerful, is also incredibly brittle.
This is generally not bad when you’re dealing with something like math because it doesn’t have to comport with experience – the axioms are true because we say they are. But when you’re reasoning about the real world it becomes a huge problem: it’s often impossible to tell if your axioms are correct. And equally difficult to know whether each of your deductions are valid.
And, if you’re reasoning in this way, if your starting axioms are wrong, or any of your deductions are invalid, all of your conclusions are totally wrong. It’s not robust to error at all. It’s a system of thought built like an upside down pyramid – the whole thing rests on one single point of balance.
Russell argues that a better way to do philosophy (or figure things out about the world in general) is to construct your pyramid right-side up. Start with known, observed facts about the world and construct your argument based on those. This will allow you to build a base of thought with strong foundations; one that’s much more robust to small errors than a purely deductive system of thought because it won’t come crashing down if one little piece is wrong.
There’s a common expression among programmers that goes, “Garbage in, garbage out.” And this is exactly how logic works. It’s a powerful tool that’s only as good as the person that uses it. And when we’re reasoning about things in the real world we want our system of thought to be like a right-side up pyramid – logic based on a robust foundation of experience – rather than a long chain of deduction that’s vulnerable to error and easy to tip over.
[^1]: If this still seems hopelessly unclear that’s because it is. To actually understand the differences you need to read some of the philosophers in question.
[^2]: he and Newton both independently came up with it around the same time
[^3]: Again this means, roughly, that it doesn’t have mass
[^4]: There’s a question about whether Bertie is really being fair to Leibniz here, but we’ll let it slide to hammer home the point.
I’ve never read one of Murakami’s novels, but I recently picked up his memoir What I Talk About When I Talk About Running. I just finished Stephen King’s On Writing, so when I spotted it in a bookstore it seemed like a logical thing to read next.
Murakami’s prose is simple and his attitude is unassuming. Basically the entire book is spent talking about running (his main pastime aside from writing). The way he writes sets up a kind of hypnotic rhythm that he builds over the course of the book. Mostly he talks about marathons: training for marathons, running in marathons, competing in triathlons, etc. But every so often the book is also interspersed with these great nuggets about writing and life.
One parallel that Murakami draws between writing and running is that both act as a sort of alchemy for the events and emotions in his life. Both writing and running allow him to take what happens to him – events, emotions, feelings – and transmute them in to another, useful, form.
“When I’m criticized unjustly (from my viewpoint, at least), or when someone I’m sure will understand me doesn’t, I go running for a little longer than usual. By running longer it’s like I can physically exhaust that portion of my discontent. It also makes me realize again how weak I am, how limited my abilities are. I become aware, physically, of these low points. And one of the results of running a little farther than usual is that I become that much stronger. If I’m angry, I direct that anger toward myself. If I have a frustrating experience, I use that to improve myself. That’s the way I’ve always lived. I quietly absorb the things I’m able to, releasing them later, and in as changed a form as possible, as part of the story line in a novel.”
One thing that I noticed reading Murakami he has very deep, zen-like thoughts on the nature of the mind and the meaning of life, but he says them so matter-of-factly that you might miss them:
“The thoughts that occur to me while I’m running are like clouds in the sky. Clouds of all different sizes. They come and they go, while the sky remains the same sky as always. The clouds are mere guests in the sky that pass away and vanish, leaving behind the sky. The sky both exists and doesn’t exist. It has substance and at the same time doesn’t. And we merely accept that vast expanse and drink it in.”
He talks a lot about things that are constantly changing in form: clouds, the sky, and the water in the river where he runs. And he uses it to place himself in the broader picture of the world:
“The surface of the water changes from day to day: the color, the shape of the waves, the speed of the current. Each season brings distinct changes to the plants and animals that surround the river. Clouds of all sizes show up and move on, and the surface of the river, lit by the sun, reflects these white shapes as they come and go, sometimes faithfully, sometimes distortedly…In the midst of this flow, I’m aware of myself as one tiny piece in the gigantic mosaic of nature. I’m just a replaceable natural phenomenon, like the water in the river that flows under the bridge to the sea.”
A lot of writers talk about pain as the engine behind the creative process. Someone like Dostoevsky comes to mind. Murakami is in this camp as well. Here he talks about emotional pain:
“As I’ve gotten older, though, I’ve gradually come to the realization that this kind of pain and hurt is a necessary part of life. If you think about it, it’s precisely because people are different from others that they’re able to create their own independent selves. Take me as an example. It’s precisely my ability to detect some aspects of a scene that other people can’t, to feel differently than others and choose words that differ from theirs, that’s allowed me to write stories that are mine alone. And because of this we have the extraordinary situation in which quite a few people read what I’ve written. So the fact that I’m me and no one else is one of my greatest assets. Emotional hurt is the price a person has to pay in order to be independent…”
Aside from just emotional pain, Murakami also talks about physical pain, which is a key part of his routine as a serious marathon runner.
“Of course it was painful, and there were times when, emotionally, I just wanted to chuck it all. But pain seems to be a precondition for this kind of sport. If pain weren’t involved, who in the world would ever go to the trouble of taking part in sports like the triathlon or the marathon, which demand such an investment of time and energy? It’s precisely because of the pain, precisely because we want to overcome that pain, that we can get the feeling, through this process, of really being alive…Your quality of experience is based not on standards such as time or ranking, but on finally awakening to an awareness of the fluidity within action itself. If things go well, that is.”
By the end of the book, he’s sort of established the rhythm of his life for us: he writes, he trains for marathons, he runs marathons. Rinse and repeat. He’s consistently doing things, but not always making visible progress.
He likens it to pouring water in to an old pan with a hole at the bottom: there’s always water running through it, but the pan never fills up. But the message he leaves us with, is that despite pouring tremendous effort in to these seemingly inefficient activities, it’s these very activities that have borne the most fruit in his life.
“Even if, seen from the outside, or from some higher vantage point, this sort of life looks pointless or futile, or even extremely inefficient, it doesn’t bother me. Maybe it’s some pointless act like, as I’ve said before, pouring water in to an old pan that has a hole in the bottom, but at least the effort you put into it remains. Whether it’s good for anything or not, cool or totally uncool, in the final analysis what’s most important is what you can’t see but can feel in your heart. To be able to grasp something of value, sometimes you have to perform seemingly inefficient acts. But even activities that appear fruitless don’t necessarily end up so. That’s the feeling I have, as someone who’s felt this, who’s experienced it.”
I majored in philosophy in college, but it can still be difficult to define exactly what it encompasses and why it’s a unique way of looking at the world.
What does it mean? How is philosophical thinking different from scientific thinking? How is it different from religious thinking? Is philosophy irrelevant in the face of our ever-expanding scientific knowledge of the world?
I recently picked up Bertrand Russell’s classic tome The History of Western Philosophy and came across this excellent discussion of the word in its introduction:
Philosophy, as I shall understand the word, is something intermediate between theology and science. Like theology, it consists of speculations on matters as to which definite knowledge has, so far, been unascertainable; but like science, it appeals to human reason rather than to authority, whether that of tradition or that of revelation.
All definite knowledge – so I should contend – belongs to science; all dogmas as to what surpasses definite knowledge belongs to theology. But between theology and science there is a No Man’s Land, exposed to attack from both sides, and this No Man’s Land is philosophy. Almost all the questions of most interest to speculative minds are such as science cannot answer, and the confident answers of theologians no longer seem so convincing as they did in former centuries.
He goes on to list a few of the kinds of questions that philosophy considers:
Is the world divided into mind and matter, and, if so, what is mind and what is matter? Is mind subject to matter, or is it possessed of independent powers? Has the universe any unity or purpose? Is it evolving towards some goal? Are there really laws of nature, or do we believe in them only because of our innate love of order? Is man what he seems to the astronomer, a tiny lump of impure carbon and water impotently crawling on a small and unimportant planet? Or is he what he appears to Hamlet? Is he perhaps both at once? Is there a way of living that is noble and another that is base, or are all ways of living merely futile? Must the good be eternal in order to deserve to be valued, or it is it worth seeking even if the universe is inexorably moving towards death? Is there such a thing as wisdom, or is what seems such merely the ultimate refinement of folly?
To such questions no answer can be found in the laboratory. Theologies have professed to give answers, all too definite; but their very definiteness causes modern minds to view them with suspicion. The studying of these questions, if not the answering of them, is the business of philosophy.
Philosophy is tasked with answering “unanswerable” questions without resorting to dogma and revealed truth. Does it really achieve that? Many times, it seems like a lot of philosophy is just concerned with debating useless stuff.
Consider the thoughts of historian Will Durant:
Some ungentle reader will check us here by informing us that philosophy is as useless as chess, as obscure as ignorance, and as stagnant as content. “There is nothing so absurd,” said Cicero, “but that it may be found in the books of the philosophers.”
Doubtless some philosophers have had all sorts of wisdom except common sense; and many a philosophic flight has been due to the elevating power of thin air. Let us resolve, on this voyage of ours, to put in only at the ports of light, to keep out of the muddy streams of metaphysics and the “many-sounding seas” of theological dispute.
But is philosophy stagnant? Science seems always to advance, while philosophy seems always to lose ground. Yet this is only because philosophy accepts the hard and hazardous task of dealing with problems not yet open to the methods of science —problems like good and evil, beauty and ugliness, order and freedom, life and death; so soon as a field of inquiry yields knowledge susceptible of exact formulation it is called science.
Every science begins as philosophy and ends as art; it arises in hypothesis and flows into achievement. Philosophy is a hypothetical interpretation of the unknown (as in metaphysics), or of the inexactly known (as in ethics or political philosophy); it is the front trench in the siege of truth. Science is the captured territory; and behind it are those secure regions in which knowledge and art build our imperfect and marvelous world….
Science tells us how to heal and how to kill; it reduces the death rate in retail and then kills us wholesale in war; but only wisdom—desire coordinated in the light of all experience— can tell us when to heal and when to kill. To observe processes and to construct means is science; to criticize and coordinate ends is philosophy: and because in these days our means and instruments have multiplied beyond our interpretation and synthesis of ideals and ends, our life is full of sound and fury, signifying nothing. For a fact is nothing except in relation to desire; it is not complete except in relation to a purpose and a whole. Science without philosophy, facts without perspective and valuation, cannot save us from havoc and despair. Science gives us knowledge, but only philosophy can give us wisdom.
Over the next few months I’m going to be experimenting with posting excerpts from books I read in this space (like this one) instead of just long essays. Ideally this will allow me write more frequently in between the times when I publish more personal long-form pieces.
When I was a kid I liked to dress up in the stories of successful people.
I read about how Bill Gates used to rock in his chair when he was thinking, and for a few months after that I made sure to rock back and forth whenever I was programming. I read a book about Stephen Hawking that described how Oxford graduate students would buy huge sketchbooks to scribble physics formulas in, and so I went out and bought a huge sketchbook to carry around with me.
We all do this. We read little snippets of people’s lives in long magazine exposés, or see adaptations of their adventures in 2-hour summer biopics, and we take their personality traits and try them on like a pair of new shoes. Then we look at ourselves in the mirror of our minds and ask: “Does this fit? Do I do that too?”
When it’s something like rocking back and forth in our chair we can usually fudge it and say, “Yeah, I do that.” Sometimes we really do end up doing those things for a couple of weeks. Something in our brains tells us if we’re similar to these people we’ll be more likely to share in their success.
But sometimes, trying on these stories doesn’t work at all. There’s a special class of stories that make our minds come up blank.
You hear a singer say, “Well, my parents always played Aretha Franklin at home so she was a big influence for me.” Or an entrepreneur say, “Well, I noticed early on that customer service was the biggest part of running a successful business, and so that’s why our vision is to help companies do better support.”
It’s like successful people have these little insightful anecdotes that they can use to explain their lives. Things like, “This has always been important to me.” Or, “I noticed this from a young age.” Or, “I’ve always been really interested in this.” These anecdotes, these calcified bits of personal history, carry a certain kind of inevitability about them. Like they were always important, and clearly pre-destined to have a huge impact on the life of the person telling them.
When we hear these things, we think to ourselves, “What did I notice early on that I can use to start my business?” or “What did my parents do that’s going to influence my future success?”
Usually, we don’t have an answer. Our brains say, “Your parents never listened to Aretha Franklin.” Or, “You never noticed that customer service was important when you were a kid.” There’s nothing obvious about our daily lives that will inevitably lead to our future success.
Why is is this? Why do we feel like we don’t have this special kind of story to tell? Why can’t we talk about these obvious parts of our lives that will inevitably lead to our success?
And, if we don’t have them, how did these successful people get them? Is it causational? Are they successful because they’ve always known what was “unique” about their experiences that they could apply to their lives? Or is something else going on? Continue Reading
- Be well read
- Learn how to build a real software business
By real business I meant a business that has paying customers from day one and that makes enough money every month that I could work on it full time when I graduated.
Although I wouldn’t go as far as to call myself well-read (I don’t think that’s something to be achieved, just pursued) I can confidently cross goal number two off of my list:
In about 11 months we’ve built Firefly up to well into the 6 figures in recurring revenue.
We’ve done this primarily through two channels: on-site website signups and large partnerships.
In July we announced a partnership with Olark to roll our cobrowsing out to their 5,000 paying customers. We’ve also struck similar deals with much larger customer support companies as well as companies in other verticals like financial services.
What makes this somewhat special is that there’s just two of us working on the business, we’re entirely bootstrapped (except for 20k from the excellent Dorm Room Fund), and this year I’m going to be a senior in college.
Because the past few years have been such a crazy ride, I wanted to take some time to write down some of the things I’ve learned along the way. Continue Reading
The other day I had coffee with two friends, we’ll call them Peter and Alex.
Peter is working on a new startup, and Alex runs a large organization. Here’s how Peter pitched what he’s working on:
Peter: So basically I’m working on a startup called W. It does X, Y and Z for large organizations. It’s especially great if you need to organize and respond to lots of support emails.
Alex: Wow, that sounds awesome. We definitely need to respond to lots of support emails for Organization X.
Peter: Yeah, it should be perfect for you guys. Would you pay for it?
Alex: Hmmm, I’m not sure. I’d really have to think about that. I think I’d have to see the product before I could say.
Here Alex is pulling away. Peter senses this and pushes further.
Peter: It could really save you a ton of time though. Think of all the hours you spend answering support emails.
Alex: Yeah, yeah definitely. It’s just I think I have to see the product itself first.
Ok, let’s stop and rewind. Where did this pitch go wrong?
The first thing to note, is that Peter is doing the right thing by asking for money.
If you’re in Peter’s position, the only way you’re really going to be able to tell if someone is going to buy your product is if they actually give you their credit card number.
But the thing to be aware of when you’re pitching your startup is that you’re not asking for an early customer’s credit card because you want their money – not really. You’re asking for for their money because it’s an excellent way to elicit useful information: people aren’t polite when money is at stake.
The problem is that when some founders ask for money, instead of using it as a tool to gather feedback, they end up steamrolling the person on the other end. They unintentionally smother the signals that are being given off in the conversation because it feels like the goal is to get the credit card.
To be clear, the goal is not to get a credit card: it’s to get information.
The first sign of trouble was right after Peter asked Alex if he would pay for his product. His response was, “I think I’d have to see the product before I could say.” With a response like that, it’s clear that Alex is pushing off a purchasing decision.
Peter sensed this, but he was in Sell Mode so he pushed further about the potential benefits of his product: “Think of all the hours you spend answering support emails.”
Now Alex feels trapped and antagonized. So instead of doing what Peter wants and agreeing to buy, he just politely pushed him off again by saying, “It’s just I think I have to see the product itself first.”
And after that second attempt the conversation was basically over. Peter learned nothing except that Alex didn’t want to buy his product that day.
But that’s not really useful information is it? What Peter needs to find out is: why didn’t Alex want to buy?
Here’s what should have happened:
Peter: So basically I’m working on a startup called W. It does X, Y and Z for large organizations. It’s especially great if you need to organize and respond to lots of support emails.
Alex: Wow, that sounds awesome. We definitely need to respond to lots of support emails for Organization X.
Peter: Yeah, it should be perfect for you guys. Would you pay for it?
Alex: Hmmm, I’m not sure. I’d really have to think about that. I think I’d have to see the product before I could say.
Here Peter encounters the same resistance as before, but instead of steamrolling over it he decides to find out where it’s coming from. The first thing to check is whether Alex can actually make a buying decision for his organization.
Peter: Gotcha, no problem. So if you were going to buy software like this, what would the process be like?
Alex: Well basically I’d take it to my team and we would discuss it. Fred is in charge of the support side of things so it would really depend on how useful he thinks it is.
Bingo. Now Peter knows that Alex was resisting because he couldn’t even buy the software if he wanted to. Now Peter should find out if there are any other reasons why Alex didn’t want to commit, and also how much his organization might be willing to spend.
Peter: Oh that makes perfect sense. Do you guys use any software that does this already?
Alex: Yeah, we actually bought Zendesk for everyone in the office. It’s about $24 bucks a month per person.
Now Peter knows that he can at least charge $24 bucks a seat. Now he needs to find out more about how they use Zendesk.
Peter: Interesting, how do you use Zendesk right now?
Alex: Glad you asked! I love Zendesk, but there are also some things I wished it had…
Do you see how easy it was to take an antagonistic pitch that lasted 4 minutes, resulted in zero sales, and generated no useful information and turn it in to a long, meaningful, and constructive conversation where both parties are engaged and interested?
Be prepared to listen when people give off subtle signals. Read between the lines. Ask for money because it’s the best way to get useful information, not because you want money.
Always try to listen more than you talk.
A few days before I have to talk in front a lot of people I get this queasy feeling in my stomach that won’t go away.
I don’t want to eat. I just want to cancel the talk because I don’t want to deal with the stress.
My hands get sweaty before I get on stage. I feel like I want to run away. I consider asking someone else to get up there instead of me.
And then I walk out there, do what I have to do, and everything works out fine. I’m usually even pretty exhilarated by the time I get done. Continue Reading
Preface: the assumption for this essay is that you’re building a B2B app, and you have something built but you’re having trouble getting people to pay for it
There are three problems with getting your first few customers:
- You (probably) don’t know how to sell things
- You don’t know who you’re selling to
- You don’t even really know what you’re selling
Nobody tells you how to answers these questions, and so most people go out to get initial traction in a haphazard way:
- They have a vague idea in mind for who wants their product
- They’ve already built the product, so they put together a landing page which which, like, totally speaks to the core value proposition
- They write some combination of any of the following:
A few half-baked ads
A few forum posts
A few comments on relevant blogs
A few blog posts
A few cold emails to journalists (because, dude, we would BLOW UP if we could just get ‘Crunched)
- They send these out into the wild, and (no surprise!), get very few responses
- They conclude that the product must suck and that nobody wants it, because Mark Zuckerberg did exactly the same thing to launch Facebook at Harvard and look at how that worked out for him
If you try to get initial traction this way, it’s very difficult to untangle why it didn’t work:
Were the ads targeted to the wrong keyword?
Was the copy not compelling enough?
Was the sample size too small?
Or does no one want what I’m selling?
When they fail to get initial traction, most people conclude that the product is the problem. That no one wants what they’re selling.
They never consider that the way they’re selling might be COMPLETELY wrong (either in the way the product is being pitched, who it’s being pitched to, or some combination of the two.)
I think most of us have been lulled into this sense that the second you post your new product to a listserve you should automatically get sucked into a 4 minute montage scene featuring dark ominous 9 Inch Nails background music only to be spat out at the end with enough money to buy that estate on the Amalfi Coast.
And when that doesn’t happen, our first response is always to blame the product.
Formally this is called the actor observer bias which tells us that we tend to blame things that don’t go right in our lives to circumstances beyond our control:
“No one responded to my emails, so the product must suck. Nobody wants it, otherwise it would already be on TechCrunch.”
This is wrong. Here’s the truth:
You have learned nothing from spending $200 on Adwords, or writing a few comments, or sending cold emails to journalists.
Let me repeat: You have learned nothing. You get a big Zero. You have no actionable information.
Your product could suck. But it could still also be completely your fault. Or it could be completely random that you didn’t get any responses. Maybe the journalists were having a bad day, or the three people who clicked on your ads were just bots.
The point is: buying a few ads, or sending a few emails, or writing a few blog posts is not enough to conclude anything.
Untangling why you’re not making sales seems like an almost insurmountable problem, especially when you realize that at the beginning you don’t even really know what you’re selling.
The problem with startups is that you have to figure out WHAT you’re selling AS you’re selling it.
It’s like having a big black bag with a product inside it, and you have to run around selling it to people you see on the street. And worse, you’re not allowed to look into the bag to know what it is you’re selling. You can put your hands into it and feel around, but that’s the extent of it.
Ok, so how do you deal with this? How do you start to figure out what you’re selling, who you’re selling to, and how to sell? How do you get those first few customers? Continue Reading