Interview with Pebble Founder: What You Might Want to Know About Hardware Companies
TechFlow Selected TechFlow Selected
Interview with Pebble Founder: What You Might Want to Know About Hardware Companies
Now, it's easier than ever to combine software engineering with hardware and launch hardware projects.
Navigating Web3 tides with focused insightsCompiled by: TechFlow
Note: This article is part of the TechFlow special series "YC Startup Class Chinese Notes" (updated daily), dedicated to collecting and organizing Chinese translations of YC courses. The twenty-third installment features the online course "Hard Tech Conversations with Eric Migicovsky," taught by YC partner and Pebble founder Eric Migicovsky.
Introduction to Pebble
Pebble was a company active in the smartwatch industry. Founded in 2012 and headquartered in Palo Alto, California, Pebble initially gained widespread attention and funding through a Kickstarter crowdfunding campaign.
The company developed a series of smartwatches known for their innovative features and long battery life. These watches used e-paper display technology and offered wireless connectivity with smartphones, allowing users to view notifications, control music playback, track health data, and more. Additionally, Pebble watches supported third-party apps, enabling users to customize them based on personal needs.
The company's first product, the Pebble Smartwatch, began shipping in 2013. It quickly captured consumer interest and achieved great success on the crowdfunding platform. The popularity of Pebble watches was partly due to their appealing design, excellent battery life, and broad compatibility with both iOS and Android systems.
However, despite being seen as a pioneer and leader in the smartwatch market, Pebble faced increasing competition as other players entered the space and large tech companies like Apple launched their own smartwatches. By late 2016, Pebble announced it would sell its core assets—including software, intellectual property, and some staff—to Fitbit.
Although no longer operating independently, Pebble’s contributions and influence on the smartwatch industry remain significant. It paved the way for future entrants and helped drive innovation across the entire sector. While the Pebble brand has exited the stage, its spirit of innovation and technological legacy continue to impact the field.
The Entrepreneurial Journey
We applied to YC in 2008 and started the company while I was still studying at Waterloo. We worked hard until 2010. Before that, we hadn't shipped anything tangible or had just begun delivering products to our first customers. One day someone told me, “Hey, you should check out YC.” So we decided to apply. We recorded our application video during beer festival season where I lived—we wore beer festival hats. We gave some wild reasons, but Paul Graham and Jessica still chose to accept us.
In winter 2011, we received support from YC—by then we may have already started shipping. Our first watch was called the "Inpulse," designed to work with BlackBerry smartphones. We had a few great customers and began generating revenue, though not much.
We borrowed $150,000 from our parents and received some grants from the Canadian government. We didn’t raise money from investors—YC was our first investor. We gradually began manufacturing watches: buying circuit boards from China, making metal casings at a small machine shop, importing plastic straps from China, and doing final assembly one by one in a garage. This early hands-on process was incredibly valuable because we could identify and fix problems in the first ten units.
During production, we encountered many issues. Our app had bugs, the watch couldn’t receive signals—it acted like a Faraday cage. So we tried replacing the metal back with plastic and used laser-cut acrylic instead. We started putting foam rubber inside the watch to prevent the back from popping off. Gradual improvements enhanced the overall quality.
When it came time to ship, we hired a few high school students and a friend to help manage logistics. We handled all these tasks ourselves in the house we all lived in, without letting operations distract us from actual sales. Looking back, I think this was important—we needed a long runway to evolve an initial idea into something truly market-fit.
The Hardest Part
We founded the company in 2008, built the initial prototype by the end of 2008–2009, and began shipping the first watches in 2010. We learned a lot from that first production run and redesigned the Pebble watch in 2012.
The process took five years. The hardest question was whether what we were doing was right—with no clear answer. I persisted, although now looking back, spending five years on it seems a bit crazy. We were naive, had limited funds, and didn’t realize building consumer hardware requires millions in capital.
I think the hardest part wasn’t external discouragement—but maintaining that naivety.
Mistakes Made
As a very small team, we had to figure things out ourselves. One mistake I made after joining YC was raising around $250,000—including a fairly large check—and using all the money to buy inventory. I thought our biggest growth constraint was unreliable stock—we didn’t have enough inventory.
Because we were assembling manually, we could only produce 10 to 50 per week. I wrongly assumed that if we could sell in real-time rather than taking pre-orders and delivering weeks later, sales would grow significantly. So we tried shifting away from hand-building in the garage and moved to a small contract manufacturer in San Jose, producing 3,000 units at once—spending all our cash on it.
But things took longer than expected—setting up a new manufacturer takes time. It took six or even eight months instead of the projected three. Costs were higher too, so our expenses increased. When we finally started production, our list of potential customers had already dried up.
Everyone switched from BlackBerry to iPhone, and we were stuck with massive inventory. That was a very difficult period.
Two Opportunities in Crowdfunding
After several cycles, I believe crowdfunding is particularly useful for two specific points, especially for startups.
First, when you have an idea but are in the early stages, you can use crowdfunding to raise funds. For example, you might raise $250,000 via crowdfunding to build a working prototype—perhaps needing just one or two renderings. This is especially helpful for software engineers trying to break into hardware.
Second, crowdfunding also works well for hardware engineers who lack sufficient funding and simply want to prove there’s interest in their product. You can email friends and ask if they’d be interested in supporting your idea. This serves as one of the best ways to test early market fit—seeing whether your inner circle is actually willing to pay for your product. It’s not perfect—some might want to try it first or hesitate—but if you’re building something people genuinely want, they’ll often fund you.
So this is an early crowdfunding opportunity. I strongly encourage people not to rush into marketing campaigns too soon. There are two reasons.
First, historically, most crowdfunding campaigns—even successful ones—fail to deliver actual products. Why enter that cycle? Actually building and delivering something to customers is highly educational. You learn what people like and dislike about your product. It shortens the time between concept and reality. And it confirms whether people are truly interested in your idea.
Second, I believe there’s another opportunity: using crowdfunding as a marketing and sales channel. I think we leveraged this more effectively within an area we’d already created. Many don’t know this, but when we launched our Kickstarter campaign, we’d already built and sold about 1,500 units—including in China.
Through that process, we went through all the hardships and build cycles, gaining strong insights into how the product should shape up—thanks to feedback from roughly 1,500 early users. They were passionate early adopters who gave valuable input, which we used to iterate and improve.
So I believe there are two distinct crowdfunding opportunities in a consumer product’s lifecycle. But many people confuse them—or don’t know which path they’re on—leading to messy outcomes.
There are also other platforms for early crowdfunding—like Kickstarter and Indiegogo—but trends show some projects are better suited for niche communities such as Hackaday forums.
If you're not confident in your capabilities, I encourage you not to rely solely on shipping-focused crowdfunding platforms—there are alternatives.
What Are the Right Requirements for Outsourcing Components?
For hard tech companies, answering this question is quite challenging because the skills required to build a product are usually highly diverse.
Within a team, you may need mechanical engineers, electrical engineers, computer engineers, and some production and operations personnel. However, you don’t always need 100% of each person’s time—sometimes just 15% or 20%, then temporarily no need for an electrical engineer. Therefore, the best solution is to have a highly flexible technical team capable of integrating everything and doing it themselves, as such teams are far more efficient than those relying on external processes.
Early in my career, I experienced painful moments when I lacked a full-time co-founder and had to work with contractors or consultants on retainer who didn’t care about the company as deeply as I did.
I mean, first, you likely don’t have much money; second, contract workers are set up differently from full-time employees who are fully committed and driving progress—they lack the same motivation and alignment.
You also face commitment issues. Just the other day, I got an email from a founder saying, “Yes, our intern went back to school,” and that intern was handling most of our business development. So in the early stage, I prefer trust, reliability, and deep domain expertise in the people I hire over mere contractual arrangements.
In short, for hard tech companies, building a flexible technical team that integrates various skills and emphasizes trust, reliability, and domain expertise is the best way forward.
Business Model Selection and Product-Market Fit
I’ve been deeply exploring this area and became very interested in Pebble. We often heard feedback like, “Can you offer a subscription?” or “Can we pay via subscription?” Everyone knows about product-market failure—the opposite of when customers are actively chasing your product.
Actually, I think what applies here is a company with the right business model. You can have the best product in the world, perfectly product-market fit. But if you choose the wrong business model, you won’t be able to sustainably invest in or grow your company.
So it’s crucial to find the right business model. Hardware or tech companies can adopt different models to drive growth. From a research and classification standpoint, some companies like Nest or Drop Cams are subscription-based, selling health ring subscriptions for $3–$10/month.
Yet others have tried similar models without success.
In our case, I don’t think subscription was inherently problematic. We tried it, but it didn’t work for us. Our mistake was failing to recognize we were in a hit-driven market.
The key is driving clicks—you launch a consumer product hoping it sparks interest and draws people in en masse.
Like game companies, we failed to realize that when developing new products and trying to create another hit, we needed to be extremely frugal with a minimal budget.
Our first product was undoubtedly a hit. I think we didn’t reinvest at the same pace into follow-up products. If we’d built a leaner organization, launched lower-cost products, tested them for success, and scaled gradually, the outcome might have been different.
For hardware companies with multiple models, it’s essential to pick the right one.
You must ask yourself: “Should we really go subscription? Is that what customers want? Will they pay $5 or $10 a month?” Forcing a model doesn’t work. Overemphasizing it might add unnecessary friction on top of an otherwise great product.
Rapid Iteration in Software vs. Challenges in Tech Companies
In software, I’m excited about shorter iteration cycles. You can push code to a repo and go live in seconds—enabling rapid product iteration and fast customer feedback. But for tech companies, it’s different.
Tech companies face different challenges: either they’re tied to hardware, which slows things down (though this can greatly shorten each cycle), or they’re developing foundational tech that unlocks new business models or use cases.
Additionally, regulatory constraints exist in areas like medical devices, drones, transportation, and autonomous vehicles. These external factors affect iteration timelines. But now, you don’t need huge upfront investment.
You can do things like startup schools—they offer free funding; for example, if you're learning AR tech, you can earn credits for it.
In medical devices, I think MVP means peer-reviewed science that supports your work. For tech companies, understanding business models, product marketing, and sales capability is critical.
Most customers get excited when they hear a solution—not just see a demo product. So communicating value and solving real problems through customer conversations is vital.
One company in YC’s demo day improved its prototype for two months before starting sales—they realized they needed to start selling and talking to customers earlier.
Another spent two years building a prototype before engaging customers. When they shifted mindset and started selling, they succeeded—landing contracts and revenue.
For tech companies, knowing how to bring a product to market and build sales channels is crucial—it may require founders to change roles and mindsets. Sales ability and effective customer communication are keys to commercial success.
Step Outside Your Comfort Zone
In the startup ecosystem, many people try entrepreneurship, but not all build profitable, sustainable companies. Especially in college towns and Silicon Valley, government grants and funding provide some safety net, but this environment doesn’t always force founders to develop products that meet real market demand and win customer support.
I believe young people should be encouraged to try diverse experiences and experiments—step outside their comfort zones. This doesn’t mean jumping blindly into the unknown, but if you’re researching a technology and believe it could positively impact your business, applying it in labs and talking to customers is important.
Attend conferences related to sales, talk to people, understand their thoughts—that’s a good start. Hopefully, you’ll find the right moment and place where a customer says: “This is exactly what we’ve been dreaming of—I can’t believe it finally emerged from the lab.” That kind of feedback should fuel your journey—but at least try, don’t stay confined to comfort.
Startup Success and Market Demand
The best companies are those that don’t require massive funding or entering a high-stakes phase. In the startup ecosystem, this applies to about 90% of companies, and under special circumstances, 5%.
Some excellent hard tech companies don’t necessarily need hundred-million-dollar investments to develop products. For instance, Oculus succeeded after 11 years of research and $2 million raised on Kickstarter—without massive backing. Similarly, another company used 3D printing to build rockets, realizing it could boost efficiency and cut costs.
In entrepreneurship, the key is understanding market demand and applying advanced tech to solve customer problems. If your product helps customers do something faster, better, or cheaper, they’ll buy it. So when choosing a tech direction, consider how it delivers real value to the market.
Ways to Acquire First Customers
A way to acquire early customers is by infiltrating communities where potential users gather—such as online discussion forums. For example, in past cases, a founder posted their app on Fire Talk forum, which drew attention and attracted the first users. Similarly, connecting with bloggers and engaging them can lead to media coverage and exposure.
In early stages, innovative promotion methods work—like attending low-cost conferences, networking with industry insiders, or even sneaking into events without tickets. By showing people the value and solution your product offers, you can spark interest and gain initial traffic.
In short, the best companies are usually those bold enough to go out, take action, and try different approaches. Doing so captures potential customers’ attention and makes them realize your product or service is exactly what they need.
Software + Hardware: A Faster Way to Start a Hardware Venture
Now, combining software engineering with hardware and launching a hardware project is easier than ever. A simple and effective method is finding an existing product and improving it appropriately.
You can directly buy products from Amazon or Alibaba—saving time, money, and tedious prototyping. By purchasing a product with half the functionality, you can receive it in two to three days and immediately start using it to solve your problem. Through real-world use and learning, you might discover your original solution doesn’t fully meet needs.
Then, you can enhance the existing product by adding better technology or software stacks. Modifications can be as simple as repurposing the same tech or changing settings. For example, scooter companies started this way—buying scooters from Alibaba and integrating phone modems and GPS.
Today, prototyping consumer and enterprise hardware is cheaper and more accessible. You can find suppliers on platforms like Alibaba, even communicate directly with factories via WeChat. Avoid hiring industrial design firms for entirely new designs—it’s expensive and risks building the wrong product. Instead, collaborate directly with factories already producing similar items and request specific improvements.
If your product targets an early market and is undergoing iteration, finding ways to reduce iteration cycles is crucial—giving you more chances to attract customers and achieve market fit.
The key is choosing the right strategy: if moving closer to the factory speeds up iteration, do it; if customer count is low, stay close to them and iterate via software. Some companies have successfully adopted this—using drones to solve specific business process issues. They moved factories into real-world scenarios—like turning a run-down motel room into a drone factory—making on-site adjustments to meet demand.
Thus, software-hardware hybrid ventures are now easier to launch—the key is selecting the right strategy for rapid iteration and market adaptation.
Intellectual Property Protection: Network Effects and Data Lock-In
Protecting intellectual property—including trademarks and patents—is a key concern. Several approaches can address this when building a company.
First, build a brand with network effects. The more people use your product, the greater your influence and recognition—making it harder for others to copy your idea.
Second, implement data lock-in strategies. By tying data tightly to your ecosystem, competitors can’t easily access valuable usage data linked to your product.
These two methods are most effective. Compared to the past, obtaining patents is harder. While some firms can help file patents, compared to the two above, patents are less accessible tools requiring extra time and money to maintain.
Importantly, if you can build a product with built-in network effects or data lock-in, you gain better IP control without relying on lawyers, governments, or courts.
Notably, few startups treat patents as a cornerstone of success—this is more common in biotech. Building a truly outstanding product is the only real way to establish dominance. Even with patents, others may still copy your idea—so patents aren’t decisive. Plus, patent strategies require money and time—hard to enforce if the company fails.
Therefore, focus IP protection on building exceptional products and leveraging network effects and data lock-in for competitive advantage.
Niche Markets and Problem Solving
In entrepreneurship, when building a hardware tech solution for long-standing problems, you may face questions. What if the problem seems too small? Actually, for investors, a small market is ideal—because it lets you clearly identify who truly needs your product and reach them easily.
Globally, only 2% care about a certain issue, while 80% do.
So for you, finding customers who are both affordable and easy to reach is crucial.
Second, small user groups give clearer product feedback. Because they know their pain point, when you say, “Buy this—it solves your problem,” they’ll use it and respond, “Yes, it solved my problem.”
That feedback is extremely valuable.
But if you target a vague, broad audience—like people without a clear product concept—it’s hard to know if someone is truly interested at any moment—maybe they lack funds.
If your target market is a fuzzy large one, failure could stem from multiple causes. So starting small is wise—but also have a roadmap to build larger products, which matters to investors.
Business Model Selection and Pricing Strategy
When choosing between cyclical models, subscription models, hardware-as-a-service, or pure product sales, several considerations arise. It’s about ensuring you capture value and align your company with the right business model trajectory.
There are many tracks available—not limited to hardware services or consumer models, but including hybrids, insurance models, ad-driven models. So it’s not binary.
At the enterprise level, calculate how much cost your product saves or revenue it generates for customers—and price accordingly. Your price should exceed their savings or profit, yet leave them with residual value, while capturing innovation gains.
Sometimes, hardware-as-a-service may be viable. This depends on your customer type and prior research.
So it’s hard to give concrete advice—more about talking to customers and experimenting. You can stress-test pricing. Most companies are small—during every PR event, new audiences hear about them for the first time and don’t know the final price. So in early stages, such experimentation is possible.
Hiring and Working with Friends: Advice
When scaling, especially in hiring, there are key considerations. Hiring is a distraction—early-stage priority is achieving product-market fit.
It’s important to clarify: hiring a team doesn’t automatically yield better results—finding a market-fit product is key.
There’s no one-size-fits-all hiring solution. General advice: hire trustworthy, reliable people—not just for domain expertise.
From my experience, when we started Kickstarter outreach, the company had just two full-timers and two interns. Suddenly, we needed to scale—spent three to six months building 30% of the team. That meant delaying initial product launch by three to six months if waiting to hire experts.
I called old college friends interested in the project and said, “Quit your job and fly over.” Within three to four days, we hired seven people—very effective. Because we were motivated, saw product-market fit, they said, “Cool, I want in.”
We avoided three-to-six-month hiring—we did it in three to six days because I knew and trusted them—they instantly joined like previous projects. In fact, your best talent pool might be people from college or past jobs.
An interesting hack: carefully review your social media accounts—Facebook or LinkedIn. Just scan names and ask: “Is this person cool?” Browsing might reveal a potential co-founder, first employee—like that long-lost high school friend who’s actually a brilliant developer.
You can hire her and bring her in. But people often miss step two: actually asking and making a real offer. Sometimes people avoid asking—afraid of rejection. I recommend a book: *The Hard Thing About Hard Things*, discussing the difficulties and benefits of working with friends.
If you end up hiring and working with a group of friends, I suggest reading that book.
How to Know When to Involve Hardware
Sometimes, certain problems can only be solved via software. But when you suspect hardware might be needed, don’t dismiss all possible hardware solutions.
I suggest starting by buying off-the-shelf devices and trying them. You might need cameras, microphones, comms platforms—just buy a Raspberry Pi and plug in USB devices. Today, you don’t need to build everything from scratch.
Some might think it’s not the right form factor or can’t do exactly what customers need.
Just build something, assemble it, and see if it at least solves the problem. Because as I often say, the initial test either proves you have a real problem worth solving—or shows nobody actually faces it—saving you three months building custom hardware.
Communicating Complex Products
For complex products—like an AI therapist or mental health coach—how do we demonstrate or convey it to users? Similar to earlier questions, we can approach this via software rather than hardware.
For such projects, I suggest first creating the assistant as an app or a 3D avatar-like model to approximate the real experience. Alternatively, consider using existing devices like HoloLens, Magic Leap, or similar—even if they cost much more than desired.
This is like early over-spending.
If possible, be willing to spend more to complete the first round of testing.
If budget-constrained, I suggest spending $400 on an Oculus device to simulate VR experiences before hardware development.
Funding and Inventory Management for Hardware Companies
Hardware company funding typically addresses R&D and inventory needs. Usually, revenue comes only after demand materializes, while manufacturing requires upfront investment and time. Long sales cycles extend cash conversion periods, making inventory management difficult.
Banks rarely lend for R&D, but family may help. Aim to minimize expenses, run hackathons, and prototype to clarify product vision.
To address this, two methods exist:
- Use pre-orders to gauge demand and fund development with deposits, reducing reliance on external capital.
- Shorten delivery timelines to reduce inventory cost and holding time.
Reducing inventory investment allows quicker conversion of cash into revenue. But sometimes higher production costs are needed to accelerate delivery. This may reduce per-unit profit but lowers upfront risk.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
深潮TechFlow
