Bridging the Valley: Developing new technologies is inherently a solution in search of a problem

Hello and welcome to Bridging the Valley, a series recording a shared exploration of the potholes in the road of hardware technology development between lab research and implementation. These obstacles often prevent the productization of new technologies, causing a phenomenon called the “Valley of Death.” In this and future articles, I dive deeper into the causes of the Valley of Death that we hypothesized in the introduction to the series, and provide more information on various areas of the technology development ecosystem.

In the startup community, it’s commonly accepted that a “solution in search of a problem” (or SISP) often results in nonviable product concepts.  A SISP is when someone comes up with a product or solution first, then goes looking for a problem for that solution to solve, or a market where that product fits. The recommended process is one where the founder identifies a problem that represents a market, then searches for an optimal solution for that problem. With this process, the founder is more likely to end up having the most effective and economical solution to the problem, and is therefore more likely to succeed in the market.  

A look at the R&D process

Unfortunately, the technology development process by nature takes the SISP approach. To explain why, let’s talk through the research and development process. It starts with an observation about how the universe functions. This is the phase generally named “basic science.”  There are no solutions or problems here — only interesting observations — but this is the foundation on which all technological development is built. 

After a particularly interesting observation is made, a researcher will say “I think we can take advantage of this to do a useful thing,” and start exploring ways to do so. At this point, no one knows what the capabilities and limitations of this new technology are — everything is speculative. The testing and analysis in the following phase of development serve to validate or invalidate the abilities of the technology to fill the needs of these speculative applications. Because the limitations of the technology are unknown, it’s impossible to have an effective problem-first approach. One can’t say that a technology is the optimal solution to a problem until its limitations are known. This means that a solutions-first approach is actually necessary for developing technologies. Thus emerges the most common explanation I’ve heard for the Valley of Death.

Understanding the problem and market

However, we shouldn’t draw the conclusion that startups emerging from the technology development process (called Deep Tech startups) are doomed to have a low success rate. Let’s take a step back. The solutions-first based approach isn’t inherently doomed to failure, although I’d concede that it does have negative impacts. The issue with the SISP approach is that it increases the likelihood that either the solution isn’t optimal for the problem or that the problem doesn’t exist at all. Fundamentally, both of these are a result of not understanding the problem and/or the market, or of being overly optimistic about the capabilities of the solution. In order to increase the likelihood of success for a Deep Tech startup, we need a supporting ecosystem that helps overcome these factors.

If we’re trying to create a system where new technologies can succeed, we want to minimize this risk as much as possible. We want the ecosystem to encourage behavior where, at the appropriate phases of development, the individuals working on the technology deeply understand the problem and the market.  

The case of federal grants

With this in mind, let’s look at the existing system. Generally, research in the basic-science phase through the early-concept-validation phase is funded by the federal or local government, academic institutions, or businesses. Of these, federal grants are the largest source, so we’ll focus on them for simplicity. 

There are numerous grant-giving government agencies, such as the Department of Defense, the National Science Foundation, and the National Institutes of Health. Each of these agencies have their own grant programs, and the specificity of the requests for grant applications vary widely. For basic science research, generally, researchers can propose their study to the relevant government agency as long as it’s within a number of very broad categories. Other grants are more specific in their requirements.  

This approach enables some of the government agencies to “request” focus in specific areas, but also enables more creativity and freedom of exploration of the natural world. Discoveries in early research phases can lead to revelations in unexpected areas, so they benefit from having this freedom for creativity and curiosity. 

Funding and market knowledge

Once a technology gets to the stage where the goal is to understand viability for application, it becomes more important to understand the requirements of the market. In general, around this stage is when a research project starts transforming into a startup. The funding at this stage often comes from either venture capital (VC) and/or a special kind of grant called a Small Business Innovation Research (SBIR) grant. In order to receive funding from either of these sources, the technology needs to have been de-risked to an adequate level, and have a compelling business plan that considers the market. 

The individuals that decide whether or not to provide the funding in both VCs and government funding agencies often have experience in industry, the necessary technical knowledge, and frequently have experience with starting companies themselves, meaning they’re adequately equipped to make a judgment on whether the application of the technology is viable. Ideally, these two sources of funding  enable the company to get the new technology to a level of maturity where it’s appetizing for a larger company to acquire or get the company to profitability.

Possible solutions and helpful programs

Although this system has the appropriate incentives by gating investments with the requirement of market understanding, it doesn’t inherently make that information easily available. Founding teams of researchers don’t necessarily know what information to look for, or where to find it. Many researchers don’t have a mind for business and are often deep specialists only in their technical field. I’ve observed a number of instances where they don’t know what to do after they’ve proven the desirability of a new technology in their lab. There are programs like the National Science Foundation’s I-Corps that help founding teams understand their markets, but these programs need to be actively sought out by the research team/founding team.  

There are many possible solutions to mitigating the SISP approach. Involving relevant industry experts with the appropriate technical backgrounds into research lab teams at established points in the research process may inject the necessary market knowledge to guide development and eliminate nonoptimal applications of the technology early. It’s also possible that setting expectations in the research community that programs such as I-Corps are necessary to proceed forward with productizing the technology may be effective. 

Perhaps commonly embedding institutions like FedTech, that connect industry serial entrepreneurs with researchers, in the research pipeline would be effective. Or perhaps engaging companies like TechLink, who help small businesses connect with relevant technologies coming out of government labs. Having market research resources available in university Technology Transfer Offices (TTO) may also help, since interfacing with the TTO is necessary when potentially beneficial discoveries are made. There are many other options, and I’d love to hear your ideas on our Discord

As usual, we need more data

We also need more information on the problem in order to choose the correct solution. There are many unknowns and assumptions I’ve made in this article, so data collection to support these assumptions is important. Data from the success rates of companies that go through programs like I-Corps can validate the hypothesis that knowledge of the market improves the chances of a successful deep tech startup. 

It would also be useful to see a study on what percentage of startup applicants get turned away from funding because they don’t exhibit an understanding of their market. This would enable us to see how much of a contributor this is. 

While we evaluate potential solutions, we also want to look at data that might support specific approaches.  For example, Harvard’s TTO, the Office of Technology Development, provides its researchers with resources for understanding how their technology may find market fit.  Looking at the success of that program will help us understand whether that strategy is effective. We’ll  look through this data in future articles.

Bridging the Valley, Hardware Handbook
Lee Wilkins

Related Posts