Where are all the hardware companies with excellent culture?
The tech world is brimming with companies that are legendary for their culture. These organizations regularly reject traditional business practices, traditional hierarchy, traditional processes. They invent new ways of doing things, new organizational models where the best builders and creators can do their best work.
All of this is with good reason: a great culture can create a wildly successful company with engaged, loyal customers. But one thing seems sorely missing in today’s pantheon of tech culture luminaries: companies that make and sell actual physical things. The only one of today’s tech giants that actually started as a hardware company — Apple — is known for its amazing products but definitely not known for its terrific culture.
So here’s my problem: I don’t build apps, I build robots.
I’m a 15 year veteran of hardware startups. I’ve worked on some really cool technology, and I’ve worked with brilliant people, but I’ve never felt that we really got the culture right. So the dialog about company culture had me really intrigued. But on closer inspection, this conversation seemed to mostly exist within software companies.
By the way, have you ever googled “hardware startup culture”? Yah, there’s a reason for that.
Hardware is Different
Part of the reason there’s so much more cultural innovation going on at software companies is just an accident of recent history. Software itself, and internet software in particular, is a young discipline. From a purely technological perspective, software can be written and released with breathtaking speed. So, in many cases, people in software have the luxury of making things up as they go along.
Hardware has a much longer history of The Way Things Are Done. You’re far more likely to find processes that exist because “this is how we’ve always solved this problem before…” Combine this attitude with long development cycles, and you can end up with a culture that can crush optimism and innovation. So maybe if we’re willing to toss old assumptions and look for new ideas, just like a software startup, we too can create an excellent, dynamic culture.
But I think there’s more to it — hardware really IS different in some fundamental ways, so hardware culture must address a subtly different set of problems.
Planning
There’s one type of company-culture blog post that drives me nuts — the post that asserts that you should just jettison any traditional planning process because it’s completely unnecessary in “the new era.” It’s usually written by someone who assumes that every single company in the world makes software and deploys it over the internet.
These posts demonize planning because it creates process. Planning begets specs, which forces you to make decisions too early, which constrains everyone’s creativity thereafter. Planning also means you need managers and more meetings. Framed like this, planning sounds so last century. Just keep working on your awesome product until it’s awesome enough and deploy it to your customers!
Well, yeah. But if you’re going to ship a physical product, you need to buy things. Like parts and subassemblies with highly variable lead times, whose price depends on the quantity you’re buying. You need to store those parts. You need actual humans to actually build your product. You need to distribute the product to retailers or to customers. All that stuff requires a lot of work by a lot of people, and a lot of coordination with other people outside your organization.
Hardware takes way MORE time, and costs way MORE money if you don’t plan.
A related concept is iteration speed — it’s a fairly well-understood principle of software design that good products are the result of lots of fast iterations, rather than lots of up-front planning. This is a key difference between hardware and software. A hardware iteration can takes weeks, if not months, and can cost thousands, if not tens or hundreds of thousands of dollars. New technologies such as (of course) 3D printing can reduce the length of the hardware cycle, but that’s still a far cry from continuous integration.
Making hardware requires a higher degree of up-front design and review because the cost of iteration, in both time and money, is higher. But that’s not an excuse to just fall back on waterfall development, and it shouldn’t be a death knell for creativity. You have to have the right amount of planning, and you have to consider how it engages and affects your people. You can use ideas like deferred commitment that grew out of software — you just have to tweak them a bit.
Structure
One key issue that comes up in the context of excellent cultures is hierarchy or the lack thereof. The plusses and minuses of a “flat” structure overall are a whole other essay, but I want to note one way in which hardware development complicates this: the sheer number of different disciplines involved.
To some degree, pure software companies have a surprisingly homogeneous problem space. Individuals on the team can contribute effectively to a large portion of the work that needs doing. If one person doesn’t deliver, another can compensate relatively easily. If a problem is more difficult than anticipated, multiple people can jump in and help out.
Building something as complex as a professional-grade 3D printer requires simultaneous innovation in software, mechanical engineering, industrial design, electrical engineering, and materials science. So you need what Ben Einstein at Bolt refers to as a full stack hardware team. These people are highly specialized — if you’re running behind on your polymer resin development, you can’t just ask a few guys from the mechanical team to pitch in.
The complexity of your team grows with the complexity of your technology.
And that’s just engineering. There’s a follow-on effect in just about every other aspect of the business as well. With a team like this, you need some structure because it helps people understand how they work and relate with each other. But you need to do this without creating the kind of silos that kill innovation at the intersections. Getting this right is critical to creating a great hardware culture.
Let’s get started
I got lucky. I found an exciting hardware company that cares deeply about its people. And it doesn’t hurt that Formlabs makes 3D printers, which according to my 10-year-old son, are almost as cool as robots.
I believe it’s possible to build excellent culture at a hardware company. At Formlabs, we seem to have managed to replicate many of the positive aspects of a modern software startup — development is fast paced and agile, working across discipline boundaries is the norm rather than the exception, we focus on results over process, we value individual autonomy over authority and control.
Our challenge now is to figure out how to scale this as we head into another inflection point in our growth (we’ll cross Dunbar’s number this year). Stay tuned — I’ll be sharing more of our progress here, and hope to start a conversation about what great hardware culture looks like.
