Hard-tech startups need to choose a business model early

When I built my first startup, we invented a technology to make applications run faster over the internet. But we were a small team of engineers and coders and the idea of manufacturing and selling network infrastructure to compete with giants like Cisco looked daunting.

So we thought that if we could license the technology to Cisco and all its competitors, that would be so much easier. We could focus on building the technology and everyone could include it in their routers.

When we approached potential partners, most of them were excited. But when we discussed license fees, we discovered they weren’t willing to pay more than it would cost to create similar functionality themselves. And worse, they didn’t understand how tricky it was to get right, so they underestimated the cost of building the functionality by a factor of three.

It didn’t take long to realize licensing wouldn’t make sense. So we bit the bullet and transformed ourselves into a network equipment manufacturer. It wasn’t easy, required a lot of money, and it took forever to convince skeptical customers that we had a better solution. But in the end, we were successful and the company was acquired by a much larger network equipment manufacturer that wanted to compete with Cisco (spoiler alert — they failed).

Now that I work with many startups, especially in the HardTech world of energy, chemicals, and materials, one of the first questions we’re always faced with is what business model to pursue — selling products directly to end-users or licensing the technology to other companies.

Direct sales and licensing are both valid business models that can lead to successful businesses. However, these two business models are very different in how they’re structured and operated, and that means they have to be funded differently, requiring a decision from a very early stage on which type of business to pursue.

Direct Sales and Licensing Business Models

Direct sales are easy to understand — the company produces some sort of product and sells it to end-users. It might be a new chemical or material, it could be a chip, a physical device, or even a piece of software. The end-user might be a consumer at home, a business, or another manufacturer in the supply chain.

In contrast, licensing is providing the know-how, intellectual property rights (patents, copyrights, and trademarks), or software to other companies that will manufacture the product themselves.

In my case, it was a choice between adding the network acceleration technology to other companies’ routers or building our own line of specialized acceleration routers. We had it easy because there were plenty of outsourcing manufacturers who could manufacture the physical products for us without us having to build and run a factory ourselves. Hard-tech companies in materials and chemistry may not have that option.

Direct Sales vs. Licensing

The attractions of licensing are obvious:

  • avoid the capital cost of building production facilities
  • avoid all the headaches of operating a manufacturing business including managing supply chains, maintaining quality control, and handling physical distribution
  • avoid having to build and manage a large sales team
  • allow the founders to continue focusing on science and technology instead of manufacturing and operations

In other words, the company can remain small and lean, consisting mostly of scientists and engineers, and focused on improving the technology and adapting it for new uses. A licensing business requires only a tiny fraction of the funding of a direct sales business and avoids all the headaches of building and managing factories to produce physical goods.

With licensing, it’s possible to build an extremely profitable business that generates great cash flow. A licensing business is far easier to build than a direct sales business, and far more likely to succeed.

Sounds great, right? Well, hang on a minute while we examine the drawbacks.

The Challenges of a Licensing Business

While the benefits of the licensing model are clear, the challenges can be less obvious:

  • Revenues are small
  • Acquisition or IPO is unlikely
  • Venture capital firms won’t fund it
  • Intellectual property rights can be difficult to enforce

Let’s start with revenues. When a company licenses its technology, the typical royalty rate is 3% of the price of the product, and rarely higher than 5%. The licensee has to build the manufacturing facility and take the risk of finding customers, and they’re only going to do it if they get to keep most of the profits. A huge success with a product generating $100 million in annual revenues means only $3 million a year in royalties to the inventor. That’s not bad, especially since it’s almost all pure profit, but not enough to get investors excited.

This leads to the second point — it’s hard for a licensing business to reach an attractive exit. The revenues are too small for an IPO. Potential acquirers have no reason to buy the business if they can get the benefits by licensing, or if the company has already made the technology widely available.

The startup could eventually be acquired by private equity which values cash flows and profits. However, PE valuations are too low to generate the huge multiples required by venture capital.

This brings up the 3rd point — without a huge exit by IPO or acquisition, venture capitalists won’t fund a licensing business. It may be a great business, especially for the founders, but it’s not a good investment for VCs. If the company needs funding, it will have to look to other sources than venture investors.

Lastly, the intellectual property rights that the company relies on its royalties can be difficult to enforce, especially for a small startup.

For the most part, big companies aren’t evil. But they are cheap. They have to be to survive. The lowest price wins customers, high profits raise share prices. They don’t set out to steal your ideas. But they have buildings full of scientists and engineers to create their own solutions.

They wouldn’t knowingly copy our technology. But if they’re approaching the same problem in the same way, it’s likely that what they come up with isn’t dissimilar from our own solution.

Maybe they were stepping on our patents. But could we be sure there wasn’t something in our product that wasn’t inadvertently stepping one of theirs? Or that there wasn’t some prior art somewhere they could dig up to invalidate our patents?

My startup had a handful of patents in network acceleration. Most of the big network equipment vendors had tens of thousands of patents that covered all aspects of networking. If we tried to assert that they were violating our patents, it was certain they would countersue, asserting our device was violating dozens of theirs.

The only way to find out if our patent held up is to take it to court. And a patent fight takes forever and is incredibly expensive. Figure on $10 million in legal fees and 10 years to work through the courts and appeals. Big companies have teams of lawyers on staff and time to wait. We didn’t.

In the end, customers want to buy products, especially if it saves money over what they’re using now. Licensing fees feel like an expense rather than an investment and nobody wants to pay them if there is a way to avoid it.

Which Business Model is Right for Your Business?

Choosing between direct sales and licensing starts with how to fund the business. If the founders can bootstrap the business without venture capital, then licensing is a viable option. If not, then direct sales are the only option.

Fortunately, a licensing business may not take much capital to build. The research might have been completed as part of a university Ph.D. program. SBIR and STTR grants may be available. Early customers may be willing to pay for pilot programs, non-recurring engineering, or discounted licensing fees to fund the MVP. Founders and employees can work for equity until the company starts generating revenue. Any remaining capital needs can be contributed by the founders or a business partner.

If licensing is a viable business model, then the choice between direct sales and licensing comes down to a few fundamental questions:

  • How willing are potential partners to license the technology vs buying a physical solution?
  • How well can intellectual property be protected?
  • Do potential licensees see the value, or will you need to create the market with end-users yourself?
  • How difficult would it be to manufacture the product yourself? Do contract manufacturers exist so you don’t have to build production facilities?
  • How attractive are the exit strategies in this space?
  • Are the founders more interested in focusing on science or building and operating a manufacturing business?
  • Most importantly: what is the preference among the founders for building a small, profitable business over which they can maintain complete control vs. the possibility of huge riches from building a large business together with financial investors?

These questions need to be answered at a very early stage before you begin fundraising. Because once you’ve taken venture investment money, you’re committed to the direct sales model.

Subscribe to receive your weekly insights