Everyone who has a success asks themselves how to repeat it, or failing that, maintain it. If you have a certain kind of career in tech you’ve seen your share of companies at this stage. Writers, musicians, and athletes struggle with it too, calling it their sophomore slump.

The very idea of luck may come from the frustration of not being able to hit that trick shot twice. ‘Lucky’ and ‘unlucky’ are backwards-looking statistics.

Or take ‘beginner’s luck’, a phenomenon where some fraction of first-timers hit a shot that they can’t replicate. It’s because they’re relaxed but also their control inputs are pretty random. Random can achieve a result that an expert would never aim for – in part because it’s not repeatable, but in part because the small successes we aim for as we grow can train us not to go for the big ones.

There’s math here. When a random process is picking winners, the only pattern to repeat wins will be that previous winners have deeper pockets (this is why people who write about trading like Nassim Taleb are so into the Kelly criterion). But we want to believe that expertise, providence, or ‘good taste’ (the combination of the two) can move the needle here.

Experienced managers have their own relationship with repeating success. For every manager who creates something from scratch there are ten more who are given something that’s not working, or working okay, and asked to fix it. The deeper thinkers of the latter category worry about whether this is something that’s possible, or if instead sophomore success is a stochastic process like beginner’s luck – except even less likely because we’re going through the motions, drilling a well in a bed that’s tapped out.

  1. Regime change
  2. Sequels
  3. Process
  4. Sustainability

Regime change

A close friend of mine grew up trading volatility in the financial crash and so did really well in a year when most people didn’t do well, but couldn’t keep it going in the vol desert that followed. Traders call this a regime change and wish they could detect them in time to switch from mean reversion to buy and hold.

Another friend operates bars in midtown manhattan and claims that tastes shift every few years, and that to survive you have to pick the next trend correctly and at the right time (too early and you lose money, too late and you lose money).

I’m on the technology side, but I watch my PM friends at companies during & after their growth snap. People try to copy themselves. Sometimes they do the opposite of what they’re doing. They define their strengths and repeat them.

People are innovative but they’re not meta-innovative – we’re good at solving problems but not picking them. Unfortunately surviving sophomore slump is ALL ABOUT picking the right problem. I’m probably a smart problem solver and I have this flaw. I spend too much time on ‘how’ and not on ‘which’, ‘why’, ‘when’. Tactics are the ‘how’, strategy and creativity live in the Ws.


For a while as a kid my favorite movie was Wonder Boys. I’ve never read the book but I love the story of how the book was written – after his first novel, Chabon got a huge advance on number two, but it was endless. He was late and lying about deadlines, and got so stressed out that he started writing a different book, one about someone in his same position. (The Phantom Tollboth kind of happened this way too but that wasn’t a second novel).

By weird coincidence, Sony brought in Michael Chabon to polish Spider-Man 2, another sequel that needed to deliver. Sequels are the best case of repeating success and they do it by manufacturing scarcity (copyright on characters and plot) and then not flubbing the execution. Tempting model here of ‘luck followed by skill’ but content creates demand for sequels in a way that other businesses can’t. And even so, it’s hard to keep clicking with audiences (search for ‘axe’ in that article).

A college classmate is a novelist as well as a professional editor and described the experience of writing her first book. “I’ll never do it that way again,” she said, I think because you know what you’re aiming at the second time through. “You write your first book front to back and every other book back to front.” I’ve never forgotten the line. Creatives are more aware of the creative process than the rest of us, naturally.

I’m not saying that I know a way to repeat success or that I know anybody who does. Often people who have repeat success are benefiting from Kelly criterion effects – the ability to invest in more projects once you have some capital. ‘Capital’ here can mean actual money, audience, installed devices – any resource you can ‘spend’ to give yourself a better shot.

Audience is a fascinating form of capital because it is directly useful for repeating yourself and tempts you to not innovate too much, to not break the mold.


I’m curious and jealous when people claim to know the process that will repeat a success. I get nervous whenever somebody wants to add process – it’s easy to claim benefits, hard to prove them one way or the other, and difficult to measure increased cost, especially at a company without experienced project managers, who are a dying breed in the age of the prod-uct manager rampant.

Most process is designed to improve work by adding conformity or authority checks. That’s good for ‘process things’ like landing airplanes and surgery, but probably less good for ‘nobody really knows’ things like exploring a problem space.

Ideally you could do meta-innovation by running a really expert series of experiments, but in the absence of an expert (I don’t know many people with this skill), your best shot is randomness. Maybe add some structure on top to make sure people aren’t taking you for a ride, but not so much that it embeds your (bad) assumptions, restricts you to hiring people you already know, and traps you at a local minimum. Maybe in the future every company will look like a VC firm.


I worry sometimes there’s no way to repeat a success other than the investment model – spending down your gains and investing in random shots.

If that’s the case, we should eat the goose that lays the golden egg. There won’t be another egg, and we have to eat. Spend the sale price on other geese. Diminishing returns is the nature of waterfowl. A golden goose is a single success. You shouldn’t double down on it, you should manage it down.

When you win the lottery or get a promotion, don’t immediately buy a bigger house. Luol Deng spent years in the same crappy house he bought as a rookie and only spent the big bucks on his pro bono soccer camp.

What’s the responsible thing for Blockbuster to do when a Netflix appears? Buy them and hobble the new, lower-margin product? Sure, if you can. And wait for another competitor to appear, better funded, able to resist your takeover attempts, and then you wind down your footprint. Keep open your profitable stores as long as you can. Return value to your shareholders and do right by your employees.

(FWIW CPG companies like P&G have been pretty good at maintaining a low cost of R&D, most of it marketing-related, and buying competitors as needed).

I’ve worked with people, smart managers in all silos, who believe in another model: eggs that lay golden geese. A process that creates repeated successes, without user lock-in, without reinvestment in variations, without wide and blind experimentation, a process that goes from ‘strength to strength’ (an absurd phrase I heard in some other kid’s report card when I was in high school – what a little suck-up that kid was).

I want to believe too.