We frequently get asked “How do you move so damn fast?” and thought we’d reveal our secrets.
What does fast mean? We’re a team with the equivalent of three engineers’ time, one data analyst, no full-time designer, and one PM. And, that PM can’t be that useful, because he’s blogging right now instead of working 😘.
We launch and analyze multiple experiments every single week — while owning infrastructure improvements, building internal tools, and prioritizing a healthy work-life balance.
So, to quote Patreon’s core behaviors, here is how we Move Fast As Hell.
Why move fast?
- The bigger the impact we want to make, the more risk we need to take
- The more risk we want to take, the cheaper each shot has to be
- If we want our shots to be cheap, we have to move fast as hell
- As a growth team, most of our experiments will return negative — so we have to move super fast to iterate our way to success.
Speed is an 80–20 thing
- 20% of the changes provide 80% of the speed gains = totally worth it
- The remaining 80% of sacrifices only brings 20% of speed gains = definitely not worth it
Sacrifices not worth making for speed
- Long hours/wellness/imbalance — it’s a marathon
- Technical debt — not real speed
- Unusable experiences — makes for meaningless experiments
Where does speed come from?
- CUT. DOWN. SCOPE. — cut things down to the simplest experiment that has any meaning. Create an “If This Works We Should” list and fill it up with ideas. This gives the same satisfaction, without the cost.
- You can stop reading here, that is the #1 thing we do. Okay, there’s also this next one:
- Individual self-sufficiency — the more steps in a project that one teammate can handle the faster things will happen. (In our case, that means engineers who scope their own projects + self-serve tools from our design and data science teams). Don’t let that blog-diva PM become the bottleneck.
- Small teams get things done — nuff said (slide 22). This is the key behind WL Gore’s massively successful engineering culture (i.e. Gore Tex).
- One clear owner — one person owns the project and owns the final calls (can be anyone on the team, as long as it’s one person).
- Avoid committee situations — Do it first, then workshop with others means come with something ready, THEN let others help you refine, rather than group design. Pro tip: Fives short 1-on-1s are better than a long meeting of five.
- Avoid “bike-shedding” — avoid unnecessary debates; when designing a nuclear power plant, don’t debate the color of the bike shed. It’s very tempting to debate details that will not make or break the project. JUST MOVE ON.
- Delegate decisions — empower people on the team to independently make decisions within their scope (“any costs under $3,000 are your call” / “design is all yours” / “you own the data analysis”)
- Enthusiasm — We’re not building features, we’re making impact. Start with the hypothesis and the history. Tie everything back to the potential.
- Bottom-up motivation — We are intrinsically motivated people. Instead of leaning on deadlines, we use dailies, weeklies, and celebrate successes and failures
Most of the places to gain on speed isn’t in engineering. It’s in the decision-making, planning, and coordinating.
Working quickly has made the growth team comfortable trying crazy ideas, and enabled us to iterate towards some large wins in Patreon’s metrics. And we have plenty of room to roller-skate even faster, with larger scissors. Terrible analogy, let me try that again.
Actually, I’m bike-shedding. It won’t make or break this post. Shipping.
Know someone whose craft would benefit from this post? Click the 💚 below to surface this post to more professionals like yourself. Cue cheesy instructional animation:
