PM Life: Becoming a Hardware PM

Rose Yao
8 min readDec 16, 2021

--

It’s almost Christmas and as a self proclaimed nerd, I have to say the best gifts come with CPUs :). There’s magic in opening the box for a beautifully designed hardware product and turning it on for the first time. I had that feeling when I got my first Apple product and I felt the same way when I bought my first Nest thermostat. So when Rishi asked me if I wanted to lead product for the Google Nest team in 2019, I said yes despite the pandemic, being a first time mom returning to work, having no hardware experience and my awesome job on Google Maps.

It’s been almost 2 years since I transitioned to a hardware PM role, and I’ve learned a ton about what it means to lead in this context. Along the way, I’ve had to break some habits from 16 years of being a software PM where the focus is on rapid iteration and experimentation, deadlines are mostly self imposed, and products are largely free.

“Hardware is hard”

We say it so often, it’s like a badge of honor. But why is hardware hard? How does it differ from pure software development?

Real Deadlines and Dollars

As a consumer software PM, I had many years of blissful ignorance of what the P&L for my product looked like. I focused on building the most engaging product, and launching when the product was ready. Even as an enterprise software PM, you have a lot of wriggle room in how you work with customers and when to “launch” a product with trusted testers, alphas, and betas.

For a hardware product, every PM has to understand the cost or BOM (bill of materials) breakdown for the product, the supply and demand curve at different price points and just how much slack there is in the schedule because the deadlines are very real.

It turns out that majority of consumer electronics sales happen in the magical months of September-December. So, if we miss that window, it has very real implications on the economics of the product line. That’s millions and millions of dollars. And once a customer makes that big purchase, they’re not going to switch or upgrade for 2–10 years. So for a hardware product manager that means there’s no lazy PMing, every milestone counts and because there are very few shortcuts. Delays can cause products to be canceled or reworked because there’s very little demand in first half of the year and the same product is unlikely to be competitive or relevant in the market the next year.

Product cycles are long and changes are exponentially expensive.

My SVP Rick gave me some great advice recently. Look at the roadmap 2 years out, if there’s any doubt, go figure it out now, because it only gets more expensive. The worst case scenario is launching a bad product.

We’re in the golden era of software development where most web products iterate daily or weekly and mobile apps are updated monthly. That means as PMs we can answer questions like is design A or B better with a 1% experiment. We can use real customer data to refine our decisions very quickly.

With hardware products, a large percentage of decisions are immutable before the product is out of the door. Worse, the cycles are long. 18–24 months is average, and some bets take 5+ years to see through (think about Apple’s M1 chips). If we make the wrong bet, for example: to change the specs of the camera module, it would cost a ~9 months delay and millions of dollars in rework. Worst, if we launch a product based on the wrong assumptions or with bad quality, the loss of trust can follow a brand for years. The nightmare example for all hardware PMs are the 2016 Samsung Galaxy phones that caught on fire.

Hardware and software teams working together

To make this all even more challenging. We just talked about how different timelines and decision-making culture are between hardware and software. But the reality is, most of the hardware we build relies on software product teams like Android, Youtube, Photos, etc to unlock the magic. Those teams have very different timelines, cultures, and constraints. There are often meetings where we are asking about 2023 which is really around the corner for the hardware team and forever for the Youtube team. It’s an interesting challenge bridging those gaps.

What I’ve learned as a hardware product leader:

Despite all the challenges I’ve listed. Year after year, fall after fall, we launch products that go under Christmas trees and hopeful create that magic moment of delight when they turn on. When I took on the job at Nest, I realized I will fail if I try to be the product leader who knew every detail about everything. There’s just years of knowledge I don’t have about everything from speaker design to display quality to camera lens tradeoffs. I’ve learned a lot about all these subjects from amazing hardware product, design, and engineering leaders, but my real lessons learned are around how I can set the right north star for the team.

Lesson 1: Define what “winning” means.

Because the stakes are high and there’s a lot of external pressure from consumers and reviewers. It’s easy for the team to chase things that don’t really matter. For example being the lowest priced product or having the best specs on every dimension. Those are dangerous strategies because they are arms races and don’t really speak to real customer problems.

So lesson 1 for me was to take a step back and define what “winning” means.

That definition should align with a real customer problem, have a sustainable business model, and we should have differentiated strategy on how we are going to solve that problem.

For example, let’s looked at the thermostats.

The customer problem is that energy in the home is expensive, complicated, and not clean. So the question was how could we create a better energy system for the home that’s affordable, simple, and clean? We looked at why customers loved the Nest thermostat. From there we decided to focus and deprioritize trying to solve for compatibility with more complex energy systems at this time. We double down on what customers loved already: iconic design and an intelligence service to save energy without compromising comfort. That led us to invest in a beautiful yet affordable thermostat and the Nest Renew service.

Lesson #2: Define quality early and often.

No one sets out to ship a bad, buggy or slow product. But it happens. On our team at Nest, the stakes are even higher, customers are paying hundreds of dollars, they will live in the home for years, and customers rely on them for everything from entertainment to energy to wifi to security.

So one of the first things I learned is that quality for hardware is particularly challenging because of opposing constraints. We want the BOM (bill of materials or cost) to be as low as possible to reach more customers, while we want the performance and features to be as high as possible to be competitive. With so much complexity how do we define what’s important and what quality means? Because if everything is important, then nothing is…

STEP 1: Define the ~3 customer user journeys (CUJs) that really matter.

These are the experiences that define the product and why customers are choosing us over a competitor. Example: the wifi in the home seamless during peak hours for the family. (kids zoom classes, parents on conf calls).

STEP 2: Translate the customer journey to KPIs (key performance indicators)

We don’t want quality to be subjective call on what’s good enough. Too many things can go wrong every step the way. So in our example what does it mean to be seamless? Define that for every team involved from the radio team to the industrial design team to the embedded software team to the mobile app engineers.

STEP 3: Align early, measure often and hold the team accountable.

The work isn’t done once we define the KPIs. KPIs needs to be aligned with engineering at the concept phase. Each KPI need to be measured every step of the way and we need to be noisy when we are missing even one critical KPI and decide what to do together.

Common pitfalls to avoid.

  • Less is more. I picked ~3 critical CUJs because reality is one CUJ probably has dozens of KPIs associated with it and even the best team can only focus on so much. Also reality is as much as product managers love to dream about features and lots of them. The best products have a single compelling reason to buy (ie: it’s the smartphone with the best camera). So be very very clear what matters and needs to be best in class and what the team should be practical about. It’s painful to realize later that a team was spread too thin and spent too much time/money/energy on features that don’t matter.
  • Don’t lower the bar. It’s easy to change the KPIs, lower the bar when we get into the development cycle. Don’t do it. My favorite hardware engineer told me it’s the product leader’s job to set the bar and his to meet it even if it takes multiple generations of the product. Ie: the ideal product is $99 with these KPIs, maybe in the first generation the price is $149 but the team knows what to aim for next.
  • Quality is everyone’s problem. Quality isn’t just on the product team. It’s not just on the engineering team. Magic only happens when everyone who touches the product from design to pm to engineering to operations to marketing to support takes ownership. It has to be an obsession for everyone. Because culture is what is rewarded and punished, as a leader I am very intentional about the incentives I set for the team.

Lesson #3: Be an optimist. Plan and dream big.

I knew I was starting to understand hardware when I wrote a 10 year vision doc about the future of the helpful home and led my team to build a 5 year roadmap. The reality is hardware leaders are the most optimistic people I know. They have to be to deal with the challenges the space presents like unexpected tariffs, supply chain issues, chip shortages, and building hardware with a remote team during a global pandemics. It can take 5, 10 years and 3+ generations of hardware to realize a vision and for consumers to really adopt the product. Even the iPhone took almost 10 years to launch and another 5+ years to reach 100M customers. It takes true dreamers to make those bets and build the products the make the vision a reality.

Here’s the secret: it turns out I already knew how to do this part of the job. I just had to give myself permission to really dream. Because the dream isn’t about the hardware KPIs, it’s the problems we want to solve for the world. For me right now the dream is simple: what it means to build a truly helpful home that released us (people like me, a working mom with 2 babies) from the endless to do lists so we can focus on what the home is really about: the people inside it.

Happy holidays everyone! Hope you have some magical boxes waiting for you under the tree. I hope some of those boxes are from Google this year ;)

Rose

Special thanks to everyone in the Nest team for teaching me all about the world of hardware. I feel so lucky that I continue to learn something new everyday from you all.

--

--

Rose Yao
Rose Yao

Written by Rose Yao

I spent the last 16+ years building products mostly at FB and Google. Also a food, travel, and fitness addict. Follow me @dozenrose or on www.roseyao.com

Responses (2)