August 2015

I recently got an email from a founder that helped me understand something important: why it's safe for startup founders to be nice people.

I grew up with a cartoon idea of a very successful businessman (in the cartoon it was always a man): a rapacious, cigar-smoking, table-thumping guy in his fifties who wins by exercising power, and isn't too fussy about how.

As I've written before, one of the things that has surprised me most about startups is how few [blocked] of the most successful founders are like that.

Maybe successful people in other industries are; I don't know; but not startup founders. [1]

An email from a founder helped me understand why it's safe for startup founders to be nice.

My cartoon idea of the successful businessman was a rapacious, cigar-smoking guy who wins by power. One surprise about startups is how few [blocked] of the best founders are like that.

An email from a founder helped me see why it's safe for startup founders to be nice people.

I knew this empirically, but I never saw the math of why till I got this founder's email.

In it he said he worried that he was fundamentally soft-hearted and tended to give away too much for free.

He thought perhaps he needed "a little dose of sociopath-ness."

I told him not to worry about it, because so long as he built something good enough to spread by word of mouth, he'd have a superlinear growth curve.

If he was bad at extracting money from people, at worst this curve would be some constant multiple less than 1 of what it might have been.

But a constant multiple of any curve is exactly the same shape.

The numbers on the Y axis are smaller, but the curve is just as steep, and when anything grows at the rate of a successful startup, the Y axis will take care of itself.

I knew this empirically, but never saw the math until this founder wrote that he was too soft-hearted, and thought he might need "a little dose of sociopath-ness."

I told him not to worry: build something good enough to spread by word of mouth and you'll have a superlinear curve. Being bad at extracting money would at worst make that curve a constant multiple less than 1 of what it might have been—and a constant multiple of any curve is exactly the same shape, just as steep.

A nice founder who extracts less just gets a constant multiple of the same growth curve—and a constant multiple of any curve is the same shape.

Some examples will make this clear.

Suppose your company is making $1000 a month now, and you've made something so great that it's growing at 5% a week.

Two years from now, you'll be making about $160k a month.

Now suppose you're so un-rapacious that you only extract half as much from your users as you could.

That means two years later you'll be making $80k a month instead of $160k.

How far behind are you?

How long will it take to catch up with where you'd have been if you were extracting every penny?

A mere 15 weeks.

After two years, the un-rapacious founder is only 3.5 months behind the rapacious one. [2]

If you're going to optimize a number, the one to choose is your growth rate [blocked].

Suppose as before that you only extract half as much from users as you could, but that you're able to grow 6% a week instead of 5%.

Now how are you doing compared to the rapacious founder after two years?

You're already ahead—$214k a month versus $160k—and pulling away fast. In another year you'll be making $4.4 million a month to the rapacious founder's $2 million.

Make $1000 a month, grow 5% a week, and in two years that's about $160k a month.

Extract only half as much and you make $80k instead. How long to catch up? A mere 15 weeks—after two years, the un-rapacious founder is just 3.5 months behind.

The number to optimize is your growth rate [blocked]. Extract half as much but grow 6% a week, and after two years you're already ahead—$214k versus $160k—pulling away to $4.4 million a month against $2 million a year later.

At 5% weekly growth, $1000/month becomes $160k in two years; extracting half as much puts you only 15 weeks behind—and growing 6% instead beats the rapacious founder outright.

Obviously one case where it would help to be rapacious is when growth depends on that.

What makes startups different is that usually it doesn't.

Startups usually win by making something so great that people recommend it to their friends.

And being rapacious not only doesn't help you do that, but probably hurts. [3]

The reason startup founders can safely be nice is that making great things is compounded, and rapacity isn't.

So if you're a founder, here's a deal you can make with yourself that will both make you happy and make your company successful.

Tell yourself you can be as nice as you want, so long as you work hard on your growth rate to compensate.

Most successful startups make that tradeoff unconsciously.

Maybe if you do it consciously you'll do it even better.

Rapacity helps when growth depends on it. Startups are different: they win by making something people recommend to friends, and being rapacious probably hurts that.

The reason startup founders can safely be nice is that making great things is compounded, and rapacity isn't.

So here's a deal: be as nice as you want, so long as you work hard on your growth rate to compensate. Most startups make that tradeoff unconsciously; maybe doing it consciously works even better.

Rapacity helps only when growth depends on it, which for startups it usually doesn't. The reason founders can safely be nice is that making great things compounds and rapacity doesn't.

Notes

[1] Many think successful startup founders are driven by money. In fact the secret weapon of the most successful founders is that they aren't. If they were, they'd have taken one of the acquisition offers that every fast-growing startup gets on the way up. What drives the most successful founders is the same thing that drives most people who make things: the company is their project.

[2] In fact since 2 ≈ 1.05 ^ 15, the un-rapacious founder is always 15 weeks behind the rapacious one.

[3] The other reason it might help to be good at squeezing money out of customers is that startups usually lose money at first, and making more per customer makes it easier to get to profitability before your initial funding runs out. But while it is very common for startups to die [blocked] from running through their initial funding and then being unable to raise more, the underlying cause is usually slow growth or excessive spending rather than insufficient effort to extract money from existing customers.

The secret weapon of the best founders is that they aren't driven by money—if they were, they'd have sold. The company is their project.

Since 2 ≈ 1.05^15, the un-rapacious founder is always 15 weeks behind.

Startups commonly die [blocked] from burning through their funding, but the cause is usually slow growth or overspending, not failing to squeeze customers.

Footnotes: the best founders aren't driven by money; the 15-week lag follows from 2 ≈ 1.05^15; and startups die from slow growth or overspending, not from failing to squeeze customers.

Thanks to Sam Altman, Harj Taggar, Jessica Livingston, and Geoff Ralston for reading drafts of this, and to Randall Bennett for being such a nice guy.

Thanks to those who read drafts, and to Randall Bennett for being such a nice guy.

Thanks to readers of drafts, and to Randall Bennett for being such a nice guy.