Reflections on the NSF iCorp program as an Industry Mentor

The following article is a reflection on my participation as an industry mentor in the National Science Foundation iCorp program. The National Science Foundation (NSF) I-Corps program prepares scientists and engineers to extend their focus beyond the university laboratory and accelerates the economic and societal benefits of NSF-funded, basic-research projects that are ready to move toward commercialization.

In a nutshell, the NSF ICorps is a 7 week program where academic teams need to interview 100 customers or partners to determine if their technology is ready for launching a scalable startup. The program is based on Steve Blank’s “Customer Development” Methodology, and guides teams to build out their own business model canvas. At the end of the 7 weeks, teams make a Go / No-Go decision on forming a startup company. I had the chance to mentor team MKCollective with Muhammad Gulzar and Miryung Kim, a startup that is focused on improving developer productivity for BigData analytics.

Team MKCollective

My favorite startup in the cohort was huMORChip - a startup that grows artificial human organs from stem cells and puts them on a testing chip for the purposes of drug discovery. Despite the fact that their startup was born out of a Michael Crichton Sci-Fi novel, they were one of the many teams that able to navigate the big bad world of pharma and identify a plausible go-to-market strategy.

My Personal Story with Customer Development

In 2012, I started a mobile application performance company called InstaOps. Because I had no idea what I was doing, I picked up a copy of Steve Blank’s Four Steps the the Epiphany and followed the customer discovery process. This included most importantly logging down each customer interview and looking for “convergence” on what features to offer. Over the discovery process, we pivoted from mobile network optimization to eventually mobile application monitoring and configuration. But the detailed logging of these interviews served another important purpose : At the time of acquisition by Apigee ( and subsequently Google), these customer interviews served as customer evidence for the value our product could bring. It was the main reason why we were able to be acquired in the first place.

However, this story doesn’t have an entirely happy ending. Because I didn’t really understand the mobile tooling ecosystem - namely the fact that Google Android and Apple could afford to give away its mobile developer tooling for free in exchange for profits from mobile ads and hardware, this essentially sets a built in expectation for mobile developer to not pay for tools. As a result, the entire market of “mobile” developer startups - Parse / Kony / PhoneGap / Appcelerator - all failed to turn into viable businesses. This could have easily been determined had we really done a close analysis of the partner ecosystem and the feedback from buyers of mobile developer startups.

What I got out of the program

1. Being exposed to a wide number of markets in a short amount of time

Because the ICorps program covers a broad range of technologies ranging from biomedical, to construction, to computer science, I received a broad perspective of the Go-To-Market challenges of each industry. Moreover, reviewing how different team’s business canvas evolved over the 7 weeks have me a unique view of how different startups approach achieving product market fit. This is unlikely to be achieved even in a venture capital firm because most VCs have a particular investment thesis and industry focus.

2. Deeper appreciation about the commercialization of applied sciences

On the final day of the iCorps program, Errol Arkilic gave a holistic overview of how science becomes commercialize - something that even I - despite being in Google’ Applied Sciences failed to understand for the last 20 years of my working life. More importantly, Errol helped illustrate the mechanics behind ”valley of death” that plagues good science making the transition to commercialization. I’m personally going to bring in some of these concepts in my work at Google Research.

3. Opportunity to reflect on past mistakes

One of the key that I did with InstaOps is failing to understand the value chain. Unfortunately, I did not recognize the the mobile software tooling market had zero profits because Apple and Google were offering tooling to developers for free by monetizing mobile hardware and ads respectively. Offering advice to MKCollective and other startups on avoiding the mistakes I made was in one word - therapeutic.

What can be done better - Enterprise Sales Training and Cost of Acquisition Analysis

NSF ICorps is a 7 week program with the ambitious task of taking inexperienced entrepreneurs and giving them the education in business. By definition, there is no way that the teams will learn all they need to succeed.

My opinion though is that the teams can benefit from doing a deep dive into enterprise sales. In my cohort, all 25 teams that participated would have needed to have an enterprise sales team to build a scalable business. Based on the presentations, most of the teams were severely underestimating the cost of acquisition and complexity of this channel.

To accommodate this activity into the course, I suggest the following:

1. Dedicate 1 of the webex sessions towards enterprise sales
2. Ask teams to interview 10 enterprise sales people of their competitors and key partners
3. Ask teams to come up with a comprehensive analysis of the productivity of their sales team, and a detailed breakdown for Customer Acquisition through this particular channel.

If the NSF team would like, I’d be happy to do a session ;-)

Final Thoughts

Coming into the the program, I thought that NSF ICorps was a paired down version of Y-Combinator for academics. However, I was very wrong - In fact the mission of ICorps is significantly more ambitious in 2 ways:

• ICorps’ focus on commercializing deep science has a much better shot at benefiting humanity that most Y-Combinator backed companies.
• The scale of the program (1543 so far) across multiple regions across america has much higher probability of spreading wealth and innovation across America

I wish the best of the NSF Foundation and I thank their service to science, America, and humanity.