In honor of the 10th anniversary of launching Twitch, I thought I’d share some of the lessons I learned along the way. Each of these insights could probably be expanded into an essay, of course. Like the ones you would want me to expand on in the future.
Most of these insights are things I heard someone else say, as a caveat. I’m not claiming that I thought this all up myself! They’re the things that I find myself telling people over and over when they ask me for advice.
Make something 10 people completely love, not something most people think is pretty good.
If your product is for consumers, either it’s a daily habit, it’s used consistently in response to an external trigger, or it’s not going to grow.
There are only five growth strategies that exist, and your product probably only fits one. Press isn’t a growth strategy, and neither is word of mouth.
The five growth strategies are high-touch sales, paid advertising, intrinsic virality, intrinsic influencer incentives (Twitch!), and platform hacks.
For internet companies, growth is more important than profit. It’s very rare for a company to achieve massive scale of use, and then die because they can’t figure out the economics. The reverse is common.
Ignore your competitors, but don’t ignore their customers.
If you’re a first time manager, you suck. That’s ok, everyone sucks. Apologize to your employees, get a coach or join a support group, read books, and generally treat management like a new important skill you can master.
Every time you add a layer of hierarchy underneath you, your job as a leader changes against and gets harder. You have to keep learning and growing. Note: good reason not to hire too fast!
You know when you need to hire: when you just can’t keep up with all the work, and desperately need someone else to take over some part of the job.
Plans are useless, but planning is essential.
Your time horizon for strategic planning should approximately be equal to the length of time your organization has existed so far.
Over time, develop a huge vision that’s bigger than any specific thing you’re working on. Put it as far in the future, and make it as huge, as you have the guts to.
You think you have a morale problem; a management problem; a recruiting problem; you don’t. You have a growth problem. Nothing succeeds like success.
Three ways to have a startup idea: something you want, something you’ve directly experienced others needing, something you’ve invented through analytic thought. They are listed in order of increasing risk.
Your culture is determined by what people perceive to be the behaviors you reward and punish. Note: Not what you actually reward and punish, and also not what you say you reward and punish.
Company cultures are reflection of their founders. To change your company's culture, seek to change how you behave. To change your company's values, seek to change what you value.
Letting an underperforming employee go is difficult and painful. You invested a lot in hiring them, and you want them to succeed. As a result you will almost always fire too late.
Presume deals won’t close and manage accordingly. Not only do deals fall through as a default, if you need the deal to close it impacts negotiations and actually makes it less likely to close.
Do the job before you hire for it. You know nothing about X, so you think you need to hire an expert in X. But you can’t tell which experts are any good until you’ve learned enough to be dangerous yourself. (Exception: cofounders)
Don’t start a company. You aren’t cut out for it. And if I can persuade you not to start a company by saying it in this tweet, definitely don’t start a company. You’re buying the economy-sized amount of effort and pain.
Today is the best time ever to start a company. You might fail, you might succeed, it’s a crazy ride either way, and you’ll learn and grow more than at any job.