Hardware startups are riskier for angels because they spend on tooling, inventory and certification before revenue, iterate in months rather than days, earn thinner margins and usually need more rounds, which dilutes early checks. They are worth it when the device unlocks recurring software or service revenue, the team has shipped hardware before, and demand is proven early.
Software founders can ship a fix in an afternoon. A hardware team that finds a flaw after tooling is cut may wait months and pay again to fix it. That single difference explains most of what makes hardware hard for an angel: every mistake costs more, takes longer to find and uses capital that a software company would never need.
Yet some of the most valuable consumer companies of the last fifteen years started as devices. Google bought Nest in 2014. Facebook bought Oculus the same year. Amazon bought Ring in 2018. Fitbit went public in 2015. I'm an angel investor in WHOOP and Eight Sleep, two hardware companies whose devices sit at the center of a membership business. The question isn't whether to avoid hardware; it's how to tell which hardware companies have earned the extra risk.
Why is hardware riskier than software for an angel?
Capital before revenue. Prototypes, molds and tooling, test equipment and the first production run all have to be paid for before the first sale ships.
Slow iteration. Hardware moves through engineering, design and production validation builds (often called EVT, DVT and PVT), and each stage takes weeks or months. Late design changes are expensive.
Manufacturing risk. Yields, component shortages and contract manufacturer problems are outside the founders' full control. The global chip shortage of 2021 and 2022 showed how a single missing part can stop a product line.
Working capital. Suppliers want payment before retailers or customers pay the company. Growth consumes cash instead of producing it, especially with inventory sitting in warehouses or retail channels.
Margins. Every unit has a bill of materials, freight, packaging, returns and warranty costs. Gross margins are typically well below those of software.
Certification. Devices with radios need FCC authorization in the US and CE marking in Europe; safety listings such as UL are often expected by retailers; medical claims bring the FDA. Each adds time and cost.
Trade and tariffs. Many devices are made in China or depend on Chinese components, and US tariffs on Chinese goods since 2018, expanded in 2025, change landed costs quickly.
For an angel, these risks add up to dilution. A company that needs more capital to reach the same milestones raises more rounds, and each round shrinks an early stake unless the angel keeps investing.
What well-known hardware failures teach
Three public stories show different ways hardware goes wrong. Pebble raised record sums on Kickstarter in 2012, proving demand for a smartwatch, but sold its assets to Fitbit in 2016 as larger rivals arrived: pre-orders prove interest, not a lasting business. Jawbone built a recognized brand in speakers and fitness bands and was liquidated in 2017 after years of competition and product problems: a brand doesn't protect a device company from bigger players. Juicero shut down in 2017 soon after reports showed its juice packs could be squeezed by hand: engineering can't rescue a product customers don't need.
None of these is a reason to avoid hardware. Each is a question to ask of the next deck: is demand lasting, can the company survive a bigger competitor, and does the device do something people would pay for without it?
When are hardware startups worth the risk?
The hardware companies that work for early investors usually have several of these traits:
Recurring revenue on top of the device. WHOOP sells memberships with the strap included, and Eight Sleep pairs its Pod with a membership for its software features. Recurring revenue smooths out the lumpy economics of selling units and makes each customer worth more over time. The story of how WHOOP built its business shows the model in detail, and WHOOP's page in my angel portfolio has the company summary.
Data that compounds. A device that collects data no one else has can improve its software with every user and become harder to copy.
Hardware as a means to a service. Robots sold as a service, sensors bundled with monitoring, or devices that enable a high-margin subscription change the business from selling boxes to selling outcomes. My guide to angel investing in robotics and automation covers how this works for machines in factories and warehouses.
A team that has shipped before. Founders who have taken a product through mass production know the suppliers, the certification path and the traps. That experience is worth more in hardware than in almost any other category.
Demand proven early. Pre-orders, paid pilots or signed purchase orders from business customers reduce the risk that inventory sits unsold.
A financing plan beyond equity. Inventory and purchase-order financing, venture debt and government grants such as the US SBIR program can fund production without diluting every round.
How to size and time a hardware check
Hardware rewards patience and reserves. A few practical habits help:
Write a smaller first check than you would for a comparable software company, and hold reserves for the next round if the company hits its milestones.
Consider joining after a working prototype and design validation, when the biggest engineering risks are retired, even at a higher price.
Ask for a milestone plan that ties each round to a manufacturing stage, so you can judge progress by what has been built.
Count on more rounds than the founders project, and model your stake after them.
For more on frontier technology with similar capital needs, read my guide on investing in deep tech startups, and for consumer devices specifically, angel investing in wearable technology. Further guides are in the angel investing section.
Questions to ask a hardware founder
What does each unit cost to make today, and what will it cost at your target volume?
What is the gross margin after freight, returns and warranty?
Who is your contract manufacturer, and which validation stage are you in?
Which certifications do you need, and when do you expect them?
What share of customers pay for the subscription or service, and how many cancel?
Where are the device and its key components made, and how would new tariffs change your costs?
Frequently Asked Questions
Q:Do hardware startups produce lower returns than software startups?
Not necessarily in the outcomes, as acquisitions like Nest and Ring show, but they usually need more capital to get there. More capital means more rounds and more dilution, so an early angel's share of a hardware win is often smaller than in a software company of the same size.
Q:What do EVT, DVT and PVT mean?
They are the standard build stages before mass production. Engineering validation tests whether the design works, design validation tests whether it holds up and passes certification, and production validation tests whether the factory can build it reliably at scale.
Q:Should an angel invest before a hardware prototype exists?
Only with strong conviction in the team, because the cost and timeline of turning a concept into a working device are hard to estimate. Many angels wait for a functional prototype and some customer validation before writing a first check.
Q:Is robotics a hardware investment?
Yes, robotics carries all of the hardware risks plus software, safety and deployment challenges. The companies that do best often sell the robot's work as a service, which turns a hardware sale into recurring revenue.