The jump from academic research to startup needs the right people

Bridging the Valley is a series recording a shared exploration of the potholes in the road of hardware technology development between lab research and implementation. These obstacles to development often prevent the productization of new technologies, causing a phenomenon called the “Valley of Death.” In this and future articles, the author will 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 this article, the hypothesized cause of the Valley of Death we investigate is that many startups emerging from academia don’t have the right people involved. Startups aren’t the only way that new technology emerging from academia is realized into society, but we’ll focus on them in this instalment and cover several other methods in future articles. Here I’ll provide a bit of background on academic research, then dive into three key questions to create a framework for evaluating our hypothesis:

1. Is “Not having the right people in the room” a real problem? Let’s sanity check our hypothesis before diving deeper. What evidence do we have that this is worth our time to explore?
2. If so, then: What tools and resources are necessary for success? In order to understand this problem, we need to first evaluate what is necessary for a successful startup emerging from academia (typically referred to as deep tech startups). Then we need to understand which of these needs are present and available to fledgling deep tech startups, and which ones are lacking.
3. How do we get the right tools for academic-based startups to succeed? What further data do we need to gather to confidently understand how much of an impact not having the right people or resources really has on technology development?

To start us off, how does academic research work? The actual research process varies enormously; every researcher I’ve talked to has a very different experience, but I’ll summarize. Generally, a university lab is run by a principal investigator (PI), who’s responsible for providing research direction to the lab and bringing in grant funding that synchronizes with the mission of the lab. The mission is influenced by the area of expertise of the PI and by statements by the major research funding entities about what they’re interested in funding. 

The actual research projects in the labs can be sparked by a grant announcement, an interesting idea of the PI, or a proposal from a PhD candidate or postdoctoral researcher. A research project needs funding for the equipment, space, and salaries involved, so the PI identifies grants that the research might fit under. A detailed research plan that outlines research goals, objectives, methods, and expected outcomes is drafted and used for applying to those grants, and guides the research if the funding is obtained. 

Once the researchers have gathered enough data to validate new technology, patents and/or academic papers are published. The resulting intellectual property (IP) is owned by the university, but if the researchers want to turn the technology into a product, they can create a corporate entity (a startup) that licenses that IP. If not, the technology transfer office (TTO) at the university will try to license that IP to companies in industry.

1) Is “not having the right people in the room” a real problem?

During my preliminary research, I reached out to a team of researchers who were developing a fascinating new method of 3D printing with very interesting new capabilities. I thought the technology showed a lot of promise, and wanted to follow them as they productized it. In one of the first discussions I had with the PI, he stated that although he wanted to see the technology released into the world, there wasn’t anyone involved that would take the torch and turn it into a startup. The PI himself was far too busy to do so.

During a later call with both him and his team, only one of the PhD candidates showed significant interest in using the technology to create a startup. The possibility of that individual doing so was discussed, but the discussion ended with the conclusion that it was in their best interest to finish their PhD in academia. Although I don’t have a large dataset, I’ve read several articles [1] [2] that also talk about this. 

I’ve also seen this become a problem in a very different phase of technology development. I recently spoke with a founding member of a deep tech energy startup that unfortunately failed. They developed the intellectual property at a university, whose entrepreneurial support system helped bring on nontechnical talent. That founder highlighted that the most likely reason for the failure of the company was that the nontechnical talent weren’t the right people. They were individuals who were used to working at large companies that required lots of administrative overhead and needed to move slowly, not in the agile manner required by startups. 

These observations support our hypothesis enough that it’s worth investigating further. 

2) What tools and resources are necessary for success? 

Given that the resources available at different universities can differ quite a bit, I’ll be making some generalizations that we’ll reassess in Phase 2. For now, let’s focus on the period of time that the technology needs to start departing from academia, right around the stage of “Concept and breadboard validation in laboratory environment“ (TRL 4 we talked about in the first article). Everything before this point is the bread and butter of academics. So what do you need to progress past this point?

1. Business acumen for building out a 10-year plan, with the ability to sell the vision of the technology to both technical and non-technical individuals
2. Legal support for IP protection and incorporation guidance
3. Enough funding for adequate runway to achieve the next milestone, including a buffer for a few mistakes along the way
4. A deep understanding of the market and its needs, along with a network into the market and the existing business structures serving and supporting that market
5. An individual with a uniting vision and interest in driving that vision forward, someone who isn’t overly intimidated by having to build their own frameworks

Let’s break them down. 

Legal support and business acumen

Many universities provide quite a lot of the resources that are needed in this area; there are a combination of TTOs, accelerator programs, venture advisory groups, and a variety of other programs that provide the expertise and guidance needed. For example, TTOs generally handle the IP, since the university normally owns the IP that was created using university tools. Accelerator programs and venture advisory groups normally provide the guidance needed to create budgets, mission statements, 10-year plans, and more. In some cases, universities even have venture groups associated with them. Legal support and business acumen is normally available, and likely not a cause.

Funding 

There are a number of approaches that can be taken, including angel/early venture funding or grant funding (including specific government grants for this stage). Realistically, this is a significant discussion in itself, so we’ll explore it more deeply in a future article.

Understanding of and network into the market 

Quite a few PIs are already very well networked into the market surrounding their field of study. However, there are many instances where the area of study isn’t necessarily the same as the market. If a lab designs a new drug, they  likely already have connections to pharmaceutical companies that would be interested in it. But if a research group designs an impact-absorption material, they won’t necessarily be connected to consumer-packaging design companies, automotive corporations, or helmet manufacturers that might benefit from this new material. In this case, there’s no reason that researchers would necessarily have connections to the market. 

As I mentioned before, there are resources like I-Corps to help educate and encourage would-be founders in networking within and understanding their target beachhead market (the customers the resulting product initially targets). I would however, consider this to be an area that’s worth exploring carefully as a potential pitfall for would-be deep tech startups. 

An individual that can drive the vision

It can be argued that the incentives in the culture of academia are misaligned with starting a company. Where net worth is often the key indicator of success in industry, the currency of success in academia is more along the lines of novel intellectual contribution, measured by papers published in esteemed journals, citations, and general recognition by peers in similar fields. Without exploring too deeply, this means that there needs to be some departure from academic culture in the values or interests of the individual driving the vision in order for them to be invested in doing so. 

Assuming that there’s someone who wants to drive the business vision, there are still other issues. At a glance, the PI would be the best person to take this role, given the leadership qualities necessary. In most cases, however, they don’t have the bandwidth to do so. Most PIs are professors and have quite a few responsibilities that prevent them from providing the level of focus required to drive a company vision forward. A PhD student would find it difficult to do this too; although it’s sometimes possible to overlap the effort of earning a PhD with the development of the company, leading the vision of a startup is a full-time job. Generally, there won’t always be someone at the lab who’s willing and able to take this role. 

3) How do we get the right tools for academic-based startups to succeed? 

Now we need to identify the right questions and start gathering data to reach an informed understanding on how much our hypothesis realistically contributes to the Valley of Death. We’ll start drawing conclusions from this data in Phase 2 of this article series, and will start proposing solutions based on those learnings.

So then, what data can we gather to support our hypotheses on the lack of our identified “needs for success”? Let’s first look at “understanding of and network into the market.” We may be able to find studies done on the causes of failure in deep tech startups, and comparing those to similar studies for non-deep-tech startups might provide some insight. Additionally, we might glean insights by comparing the success rates of startups emerging from I-Corps (or similar programs) with startups that didn’t participate in those programs.

Determining how often the lack of a vision-driver causes technology development to stall is more difficult. I’m aware of studies on motivation of researchers to commercialize technology, but not of analyses of the impact from lack of bandwidth. I’m open to suggestions for data that could indicate the impact of this hypothesis.

Although we determined earlier in this article that it’s unlikely that lack of legal support and business acumen are contributors to the Valley of Death, it’s based on the assumption that my exposure to universities is a representative sample. Most of the universities I’ve interacted with have been entrepreneurship-focused, like MIT and Stanford.

To confirm that our conclusion isn’t incorrectly biased, we should look at data around what percentage of research universities actually have these resources available to aspiring entrepreneurs. I’d also like to confirm that these resources are actually effective. Business acumen isn’t something that can simply be taught (it’s tacit rather than explicit knowledge), and good advisors are absolutely helpful, but only effective if called upon at the right times. I hope the “causes of failure in deep tech startups” studies will also provide insight into this.

These questions should be adequate to evaluate the cause we’ve focused on in this article, but they won’t capture other causes that we haven’t thought of. In addition to these questions, we also need to continue looking for data on the other issues along the path of technology development, including listening to the stories of those who’ve experienced the Valley of Death. If you have stories that you’d like to share or suggestions on what data we should be looking at to verify our hypotheses, I’d love to hear about them on our Discord server!