How to Build a Company around Crafters 🔨

Interview with Kaz Nejatian, COO & VP of Product at Shopify.

Sep 7, 2023
5

Kaz Nejatan is a lawyer by training, and, early in his career, he would spend most of his time in search of meeting rooms. This is literally the first thing he told me last week as we sat down, precisely, to talk about meetings.

That’s because, as COO and VP of Product at Shopify, he hit the headlines earlier this year for abruptly deleting 12,000+ meetings from the calendars of its employees.

Abruptly, here, means he just ran a script.

At some point between Christmas and New Year’s Eve, Kaz and Tobi Lutke, CEO, snapped their Thanos’ fingers, and all recurring meetings with more than three participants were gone.

I did my homework for this chat, so I already knew the story. Entering the call, I felt prepared. I had a lot of questions: what’s the difference between good and bad meetings, how do you foster async/written communication in engineering teams, how do you make remote work effective, and more.

The interview, though, turned out to be about much more.

Shopify has a unique culture. You can’t understand how and why they removed meetings, without first understanding what Shopify believes about the world: what they cherish, what they fear, and the extreme lengths they would go to protect the former and prevent the latter.

It was an incredible chat, and here is what we will cover today:

  • 🐒 Chaos Monkey — the raw numbers, before and after, of the operation that purged 12,000+ meetings.

  • 🔥 Chaotic and different — how Shopify protects its culture from scale and bureaucracy.

  • ⚔️ Crafters vs managers — the key conflict around which companies are built.

  • 🔨 Empowering crafters — organizing a company around people who build.

  • 📌 Bottom line — what makes Shopify truly unique and what we can learn from it.

Let’s dive in!


🐒 Chaos Monkey

🔗 Video Chapter

One of the key themes of Refactoring is how to make product/engineering teams productive.

This is always hard to address, because different people have different needs, partly based on their jobs — e.g. engineers vs managers.

So, one of the most debated and divisive topics in this space is, of course, meetings. I believe for two reasons:

  1. Everybody has meetings, so everybody has opinions on them.

  2. Meetings are often held for the lazy convenience of some of its parties (e.g. managers) at the expense of others (e.g. engineers).

Most companies these days agree about the need for reducing meetings, but it’s rare to see big tech actually walk the talk. Shopify, instead, in a few days and without notice, automatically deleted all recurring meetings with more than three people, from everyone’s calendars.

They dubbed this move Chaos Monkey — a nod to Netflix’s own Chaos Monkey: a tool that tests infrastructure resilience by randomly shutting down parts of it. Shopify’s monkey did other things, too, like deleting various Slack channels, but the meetings angle is what got most of the headlines.

But what were those meetings? Kaz says that the first, very practical effect of this, was banning all-hands.

At Shopify this is a big deal: each person typically belongs to multiple teams, some of which are cross-functional and vertical to some product area, while others are about some horizontal craft. There is flexibility around how teams are managed, but typically most of them have a weekly all-hands.

Kaz calculated that the average person was spending about 6 hours/week just in all-hands. It felt like things were just getting out of… hand, if you see what I mean.

Fast forward 6 months, as the dust settled, here are the results of this operation:

  • 📉 -16% meetings — on average, during the week.

  • 📈 +18% productivity — based on projects shipped this year, thanks to faster turnaround in almost every part of the production cycle, from specs to PRs.

Now, in most companies — especially big, public orgs like Shopify, with 10K+ headcount — such an initiative would have involved roundtables, change management plans, initial tests on a few teams, and plenty of measurements, before rolling it out to the whole workforce.

Instead, it was done in just a few days, under festivities, and almost without notice.

How is that possible?

To understand how Shopify can afford to operate at the speed of thought of a small startup 1/100 its size, we have to understand how Shopify thinks about itself in the first place.

💥 Chaotic and different

Shopify’s customers are entrepreneurs, creators, and people who are successful thanks to their own ingenuity and resourcefulness.

Try to picture that.

Shopify literally lives and breathes entrepreneurship. It wins by empowering people with tools that support those qualities, and, for most of its years (it was founded in 2006), it survived as an underdog, in a world of e-commerce behemoths, by displaying the same qualities of its customers: creativity, speed, flexibility.

Kaz and Tobi know that, and as the company grows larger and larger, they hold on to the culture that made it special. More often than not, this means embracing being different, and a bit chaotic.

1) Different 🔀

🔗 Video Chapter

Shopify has a bias for being different. Where most companies zig, it looks for opportunities to zag.

Shopify is remote-first since long before covid, they built their own version of JIRA, called GSD (Get Shit Done), and have a lot of custom tooling for which they could otherwise use off-the-shelf products.

This is, of course, both an opportunity and a risk. Custom tools often turn into nightmares to maintain, and I personally do not recommend startups build their own plumbing unless they are core to the business.

But the argument here, by Kaz, is that such tools are in fact core to Shopify. Shopify is successful because it is different, and to be different you need different tools.

Different doesn’t always mean better, and you will definitely make mistakes, but as long as you admit them and recover fast, such a bias is a net positive for the company.

But how do you stay different when you are a public company with more than 10,000 employees?

It turns out, by embracing some chaos.

2) Chaotic 🔥

🔗 Video Chapter

Kaz is not a man who steps away from controversy. He tells me most big companies “are fucked” and “usually have no idea of what they’re doing”, so you shouldn’t copy them.

To avoid turning into one of them, Shopify moves at the speed of a team 1/100 its size, and routinely accepts some level of internal chaos, trusting that people will eventually figure out.

There is an emblematic story about this.

When the team was still in the hundreds, and before it was remote-first, they had to move from their office to a new, larger one. The lease for the old office, though, expired one month before the new one started, so there would be this one-month hiatus where no office was available.

Now, 9 out of 10 CEOs of what was already a sizable company would have probably just extended the current lease by a few months, not only to cover the gap, but also to create a comfortable cushion during which both offices would be available.

Tobi said: “screw that, for one month everyone will just figure out where to work from”.

It worked. No big drama. And Kaz thinks that set the foundations for eventually making the company remote-first (or, as they say, digital-first).

But why do other companies fail at this? Can’t everybody just work the same way? Kaz has a theory.

⚔️ Crafters vs managers

Kaz believes that at the heart of the issues that plague most large companies, there is a conflict between what Shopify calls crafters (e.g. engineers, designers) and managers.

In any tech company there are way more crafters than managers, yet systems, schedules, and big decisions are mostly dictated by managers. The problem with this is that crafters and managers do extremely different work, and get energy from different things — see PG’s essay, Maker’s Schedule, Manager’s Schedule — so optimizing systems for managers makes good crafting eventually die.

Kaz thinks this is a historical problem: your average company structure today is very hierarchical — and management and org structure ideas are typically from decades ago, and are just suboptimal for today’s tech teams.

Kaz does not discount the crucial role of managers, but he argues that companies should be organized around crafters’ productivity, not the other way around.

Conversely, when companies get organized around managers, they eventually produce career ladders instead of actual products. People play promotion games, politics, and try to game the system to rise the ranks, because that’s where the incentives are.

And that’s how great products and companies die.

Shopify, instead, wants to build a company around crafters. But how do you do that? Out of everything that Kaz told me about this, three things stuck with me the most:

1) Separate career tracks 🐾

Shopify has dual career tracks for managers and ICs. This is a reality in many companies today, but Shopify is adamant about this in two main ways:

  1. Becoming a manager is never a promotion, but always a step to the side on the ladder. It’s a different job, and the company is crystal clear about it.

  2. Crafters and managers have exactly the same opportunities in terms of career advancement and compensation. This is not trivial: many companies sport dual tracks but managers still get to earn and advance more than ICs.

You can find more advice on creating good career frameworks in this two-part series 👇

2) Default to no ❌

When your manager asks you for a meeting, Shopify encourages that your default answer should be no, unless you believe you can have it without losing productivity.

Conversely, managers know they should get information primarily through systems rather than people, so, when something’s missing, the canonical solution should be to make the systems better. Meetings are the last resort.

About this, Kaz tells me that Shopify treats meetings like a missing API: you won’t have an API for everything at any given time, and it’s ok, but you have to know that that’s what you should tend towards.

I published a framework and a case study for reducing meetings in a previous article 👇

3) Remove things ✂️

Across several experiments, researchers demonstrated that, as humans, we have a bias towards addressing problems by adding new things (e.g. components, initiatives, ideas) rather than removing existing ones. This is unfortunate, as subtraction is often faster and cheaper than addition 👇

Shopify, instead, has a bias towards removing things, which, coupled with the chaotic vibes, means it is not afraid to burn things down before building them back up.

📌 Bottom line

Shopify is a big tech that is terrified of becoming a big tech.

More than any other company of their scale that I know of, they go to extraordinary lengths to retain what made their culture special as a startup. This clashes with their size, sometimes, but it’s a calculated risk.

People are more resilient than we think — they can handle a little bit of chaos. Eventually they just figure out” — Kaz Nejatian

This insight by Kaz ☝️ is key to understanding how Shopify thinks of itself.

But I can’t help but think that you can’t pull this off unless that’s how you have always operated: you have always hired fast-movers, and you have always focused on crafters.

Kaz studied as a lawyer, but he writes code almost everyday. Tobi is the CEO and pushes to production weekly.

You can’t just retrofit this.

Shopify is unique in how it takes the hustle, bias for action, and intensity of classic startups, and blends it with the intentionality, systems, and care for crafting of async, full-remote companies.

By looking at most examples in our industry, we may have lived under the impression that these qualities are incompatible. You are either an Elon company, or you are a GitLab, or a Doist.

Shopify says there is a third way — but brace yourself, it will be a ride!


And that’s it for today! If you are finding this newsletter valuable, consider doing any of these:

1) ✉️ Subscribe to the newsletter — if you aren’t already, consider becoming a paid subscriber. You can learn more about the benefits of the paid plan here.

Get full access to Refactoring ✨

2) 🍻 Read with your friends — Refactoring lives thanks to word of mouth. Share the article with your with someone who would like it, and get a free membership through the new referral program.

Read with your friends 🍻

I wish you a great week! ☀️

Luca