What I’ve learned from the companies I’ve worked at

In my ~12 years of experience in the tech industry, I’ve worked for 4 companies (not counting my current role): Amazon, Snap, Google, and Discord.

Here’s a summary of one thing I learned from each company.

First, some overall lessons:

1. It’s all about the team. Being on a great team defines most of the experience at a company. Even at a well-run, growing company, if you’re on a team that’s not prioritized or poorly managed, you’re going to be pretty limited. I’ve seen both sides of this, so I truly understand the contrast. Unfortunately in my case, I have more experiences being on the bad team than the good one.
2. Most companies are dysfunctional in multiple ways. And the larger the company gets, the more vectors there are for mess.
3. As you get more senior, there’s less incentive to rock the boat — even if the boat really needs rocking. There was a time at Discord that I tried and failed to solve a VP-level issue with another team. It was such a glaringly obvious problem yet none of my leaders wanted to get close to it. I realized that in hindsight there’s little incentive for Execs to go out on a limb to solve tough problems that they don’t fully control.
4. Growth lifts all boats. The best place to be is at a company that’s growing quickly. People are happier. There’s less clinging and power moves. The inverse is true as well. From my experience, post-growth or stalled growth are worst phases for a company. As an employee in this phase, your expectations continually have to be reset by reality. I’ve been this stage at two different companies, and it’s not fun.
5. The product function shifts from building to communicating/coordinating. As a company scales, the PM’s role becomes more about managing information/perception and internal marketing — and less about actual building.

So, here’s the most important lesson I took from each of these companies:

What I learned from Amazon: Clarity of written communication

Amazon’s 6-pager process is special. I wrote several of them in my time there, and even had to present two to Jeff B himself.

The amount of context and information that goes into one of these documents is staggering. It’s obvious why Amazon uses them over powerpoint presentations, which are so easy to fake by comparison.

Writing the 6-pager taught me to go extremely deep on a topic and understand multiple angles around it. In fact, I don’t think I’ve had to get to that level of depth on a topic since then!

What I learned from Snapchat: the power of visual communication

Snap is a design-driven company, so most ideas are communicated visually. I wasn’t very skilled in visual communication when I started, so it was a struggle for me at first to distill LOTS of information into a prototype or an illustration on one slide.

But I got lots of reps with and started to enjoy the process.

The culture of visual communication also spilled into the product itself. I was ingrained with pixel-level attention to detail after having to test and review so many features over the years.

It has become abundantly clear to me in the years since I was at Snap that visual communication is a skill that doesn’t come naturally to most PMs, and most of us aren’t great at it (at topic for another post).

What I learned from Discord: design debt is as bad if not worse than tech debt.

All companies fear tech debt. At Snap, the Android app was ignored for years. It wasn’t until it started to impact user growth — most of the world outside the US are Android-heavy markets — the the problem became too big to ignore. Eventually, the company had to rewrite the entire Android application from the ground up. It was a very, very costly project that took several years.

So I get tech debt. But Discord introduced me to design debt. This is when the an engineering-driven company grows fast and keeps adding feature after feature without much design direction.

Discord emerged from a gaming studio. Today, the product itself resembles a game: there are tons of complex interactions and interfaces that are easy for power users who have “leveled up” but are pure hieroglyphics for new users. This design challenge for new users is what I think makes it hard for them to retain at higher levels.

Just like tech debt, design debt is debilating and is extremely costly to fix. I would argue that it’s as costly to fix for product-dense companies.

What I learned from my time as a founder of a venture-backed startup

I’ve actually already written about that here!