What Makes HardTech So Challenging?

The primary reason few hardtech startups have reached unicorn status is simple: launching a hardtech venture is incredibly demanding. Software is a venture capitalist's dream—write a check, make a few connections, and wait for a big exit in five years.

With just a handful of developers, a software prototype can be created during their free time, often utilizing free resources. If specialized skills are necessary, they can easily be found on platforms like Upwork.

Before seeking funding, a software startup may already have a minimum viable product (MVP) and even some paying customers. A few hundred thousand dollars can help kickstart marketing, followed by a couple of million to scale operations. It’s straightforward—minimal capital requirements can lead to enormous growth.

Software products are infinitely scalable. Connecting to AWS can take a user base from ten to ten million in no time, without the headaches of physical production. The necessary skill sets are widely available, and teams can work remotely from anywhere.

When challenges arise, solutions are often just a search away on Stack Overflow. If you encounter a problem, switching tools or tweaking source code can often resolve it. But how do you approach issues when developing a new lithium-ion cathode?

Creating a superior battery electrolyte or a plastic recycling method requires specialists, lab facilities, and capital for inventory, not to mention the need for a physical factory. Have you ever built one? I certainly haven’t.

While establishing a software startup is no cakewalk—90% still fail—a hardtech venture is exponentially more difficult. If software is a dream for venture capitalists, hardtech is often a nightmare. It demands specialized skills, substantial capital just to validate technology, and investors who comprehend the intricacies of tech rather than mere spreadsheets.

A HardTech CEO's Daily Realities

Much of my career has been in two computer networking startups. Although we aimed to be software-centric and considered ourselves software developers, our product was ultimately hardware that customers needed to install.

I assumed it would be straightforward; we didn’t have to manufacture ourselves, as there are many electronics contract manufacturers, from small local shops to major factories like Flextronics and Foxconn.

Since our devices utilized standard components with minor adjustments, how complicated could it be? I thought I could focus on corporate strategy, customer relations, and marketing.

Instead, I found myself grappling with hardware issues. I was pleading with manufacturers to produce the necessary hardware in quantities that were trivial compared to what Apple or Cisco required, while simultaneously trying to secure enough cash and space for a year's worth of inventory—the minimum they would produce for us.

When the inventory finally arrived, I believed I could concentrate on sales. But soon, customers reported that some units wouldn’t boot. It turned out the manufacturer had used expired motherboard batteries. Some malfunctioned within days, while others lasted years. This led to a global recall of hardware already installed in customer networks.

Even in smoother times, my days were spent apologizing to clients for lost shipments, navigating regulatory certifications, and dealing with customs delays.

During one particularly challenging holiday season, I was in Taipei trying to troubleshoot faulty electronics shipped to customers. The issue stemmed from a seemingly trivial mistake—a chip designed for PCIe v2 instead of v3. It functioned inconsistently, and I lost months of shipments and a significant portion of my sanity over this oversight. The only thing that kept me out of the hospital was strong medication.

My most frustrating experience involved a local electronics firm that went bankrupt just as we began shipping our hardware. We had to secure a court order to reclaim our half-finished products from their facility.

Can you imagine the chaos? I often found myself screaming about everything from sales tax audits to faulty hardware and lost shipments. Every hardtech CEO needs a private office—not for quiet reflection, but for venting frustration.

Yet, even these complications were minor compared to the challenges faced in true hardtech—developing new materials, chemistries, and hardware systems.

Why Not License Technology?

I frequently work with early-stage hardtech startups founded by graduate students aiming to commercialize their academic inventions. One common strategy to mitigate the costs and complexities of large-scale production is to license their technology to established companies.

While it appears straightforward—companies like 3M or BASF could pay a portion of their product costs to manufacture and sell the product themselves, allowing founders to focus on R&D—it comes with challenges.

Firstly, large companies are often more inclined to develop their own technology rather than license it, and the licensing fees are typically much lower than anticipated (1-3% is common).

Secondly, while licensing can bootstrap a business, it’s not ideal for venture capital investment. Licensing models may yield substantial profits but don’t usually lead to large exits, often resulting in private equity that pays only a small multiple of EBITDA.

If you’ve developed great technology, you’ll need to scale it independently. You’ll have to navigate manufacturing, supply chains, shipping, and sales to industrial clients. If research is your priority, hardtech may not be the right field for you; quality control often takes precedence over research and development.

Toll manufacturers can assist in running a factory and handling logistics, which can be beneficial at the outset. However, it remains your product and factory that you must oversee, treating it like a rented facility rather than fully outsourcing production.

HardTech's Challenges

If you’re determined to start a hardtech venture, be ready to tackle these challenges from the onset:

1. The capital needed for production facilities and inventory vastly exceeds that of software.
2. The time to move from prototype to production is significantly longer than for software products.
3. Software bugs can be resolved with patches in days, while even minor issues in chemical production can be catastrophic.
4. Founders are often scientists lacking business expertise. Even with a business partner, the focus may be more on sales and marketing than on manufacturing and supply chain management.
5. Acquisitions typically offer lower multiples based on EBITDA instead of high revenue multiples.

In summary, hardtech demands more resources, takes longer, has a higher failure risk, and provides lower returns on success. It’s no wonder venture capitalists are hesitant.

Bringing these factors together creates a formidable challenge: transitioning from a significant invention to profitable production and, ultimately, a successful exit.

The Imperative to Persist

With all these hurdles, is there a reason to pursue a hardtech startup? Absolutely! These innovations are crucial and urgently needed to address our climate crisis and build a sustainable future.

While software has its place, there's only so much time we can dedicate to new social media platforms or meal delivery applications. If we invested even a fraction of our resources into solving energy storage issues rather than creating cryptocurrencies, we might already be on the path to carbon neutrality.

Building a hardtech startup may seem nearly impossible, but I urge you to take the plunge. The world desperately requires tangible solutions to pressing challenges, far more than another software application.