Group of entrepreneurs surround a computer screen while looking at software.

Episode 3: Startups from university software | Predicting Startup Success podcast

18 November 2026

The article at a glance

In Episode 3 of the Predicting Startup Success podcast, Chris Coleridge speaks to Taylor Bench about a different kind of prediction challenge: identifying university-developed software with commercial potential before a startup, founding team or business model even exists.

Taylor Bench.
Taylor Bench, Managing Partner, Summit Venture Studio

Sometimes a startup begins with an entrepreneur searching for a problem. Sometimes it happens the other way around. 

One of the examples discussed in Episode 3 of the Predicting Startup Success podcast came from Duke University Hospital, where an anaesthesiology department developed software to solve a scheduling problem. The software worked so well that it generated around $6 million in annual savings before anyone seriously considered whether it could become a company. 

That story captures the central question explored by Taylor Bench, Co-Founder of Summit Venture Studio, in conversation with Chris Coleridge. 

The problem was real and the technology already existed. What was missing was the company. 

As Coleridge puts it “The challenge is not, is this a good startup? Because the startup isn’t there yet.”

For investors, founders and universities, this creates a very different prediction problem. Instead of evaluating an existing venture, the challenge is deciding whether a technology could become a successful company if the right people, processes and commercial expertise were built around it. 

The challenge is not, is this a good startup? Because the startup isn't there yet.

Chris Coleridge

Why software often falls through the cracks 

Universities create an enormous range of commercially relevant technologies. Alongside patents and scientific discoveries, they produce clinical tools, data infrastructure, workflow systems, teaching platforms and countless pieces of software developed in the course of research. 

Yet much of that software never reaches the market. 

According to Bench, one reason is historical. Technology transfer offices have traditionally focused much of their attention on patented intellectual property, while software has often received less scrutiny and fewer resources. 

“Most of the attention would go towards patented IP versus copyright or curriculum or software.”

The issue is not simply one of attention, but also one of expertise. 

Bench recalls that when he worked in university technology transfer, he was often the only person in the office with experience of writing code or working in software businesses. Many colleagues came from legal or scientific backgrounds, bringing deep expertise in patents and research, but less familiarity with software commercialisation. 

That picture is starting to change. Universities are increasingly paying attention to software opportunities because they can often reach the market more quickly, with lower costs and fewer regulatory hurdles than many traditional deep-tech inventions. 

But the episode makes clear that a significant opportunity remains hidden within university portfolios: technologies that solve important problems but have never been developed with commercialisation in mind. 

Building companies, not just finding them 

This is where the venture studio model differs from more familiar approaches to entrepreneurship. 

Venture capital firms typically invest in existing startups. Accelerators support founders who have already decided to build a company. Technology transfer offices help manage and license university intellectual property. 

Summit Venture Studio operates at an earlier stage. Rather than looking for complete startups, the team looks for technologies that might become startups. 

That process begins with screening hundreds of opportunities every month. Summit has developed internal AI tools that assess technologies against a range of commercial criteria and benchmark them against thousands of university disclosures. 

The goal is not simply to rank opportunities but to identify what is missing. 

Who is the customer? What alternatives already exist? Is there evidence of market demand? What information needs to be gathered before a decision can be made? 

And importantly, who should lead the company? 

For Bench, recruiting the right CEO is one of the most important parts of venture building. He thinks the ideal candidate is rarely the inventor. 

Instead, Summit typically looks for experienced industry operators who understand the market, know the customers and are capable of leading sales. What matters most is not ownership of the idea but a commitment to the problem. 

As Bench explains: “I’ve spent years in this industry, nobody’s solved this problem, and now here’s a solution. I want to take that to market.”

That insight challenges a common assumption in entrepreneurship that founders are most motivated when commercialising their own inventions. In the world of university software, conviction often comes from years of encountering the problem rather than creating the solution. 

I've spent years in this industry, nobody's solved this problem, and now here's a solution. I want to take that to market.

Taylor Bench

The messy reality of turning technology into a business 

While the venture studio model relies heavily on frameworks, scorecards and playbooks, the conversation repeatedly returns to the limits of process. 

Bench clearly believes in structure – Summit uses benchmarking, validation frameworks, customer discovery programmes and internal scoring systems to evaluate opportunities. The studio has even analysed more than 15,000 university disclosures to improve consistency in decision-making. 

Yet he is equally clear that uncertainty never disappears – once a technology enters the market, “There are so many variables,” Bench says. 

For example, different industries buy differently. Different customers respond to different messages. Marketing tactics that work for one company may completely fail for another. 

The challenge is therefore not to eliminate uncertainty but to learn as quickly as possible. 

That learning takes place through customer interviews, validation boot camps, conferences, webinars, landing pages, sales conversations and product testing. Throughout the discussion, Bench returns to a simple principle: every assumption should be challenged as early as possible. 

In that sense, the venture studio’s role is not just to build companies, but to build learning systems around them, translating the knowledge into impact. 

Towards the end of the episode, the discussion turns to one of the less visible obstacles facing university entrepreneurship. 

Commercialising academic technologies means navigating institutions with multiple stakeholders, different incentives and often complex ownership structures. 

Bench jokes that describing these experiences as ‘war stories” is optimistic – ‘Current trench stories’ might be more accurate. 

Universities differ in how they organise research, manage intellectual property and make decisions about commercialisation. Some technologies emerge from individual inventors. Others have dozens of contributors spread across departments and projects. 

For many investors, that complexity acts as a deterrent. 

Bench recalls speaking with a venture capitalist who told him that after completing one university deal, he never wanted to do another. The process was simply too difficult. 

This complexity helps explain why specialist venture studios exist in the first place. Part of their value lies not only in identifying promising technologies, but also in navigating the institutional landscape required to bring them into the market. 

Yet even after the licensing agreements are signed and the company is formed, uncertainty remains.  As Bench puts it: “The real predictor is when you actually start executing on something and you see what the market says.” 

This article was published on

18 November 2026.