Start Your Startup with the Customer’s Needs

Reporter interviewingMillions of dollars of U.S. taxpayer money are funneled into a wide range of startups every year. Any taxpayer might reasonably ask, what do we get for our money? Quite a lot, according to the U.S. National Science Foundation (NSF): “Some well-known firms, such as Qualcomm and Symantec, received early support from the NSF program. Between 2014 and 2018, firms funded (or previously funded) by our program raised more than $6 billion in follow-on private capital and reported over 90 successful exits (IPOs, mergers, and acquisitions)…”[1] See their Showcase page for examples.

But more startups should be succeeding, and the problem was not lack of effort or bad research, NSF decided: It was the failure to learn what customers were actually willing to pay for, and how they would pay for it. Partly for that reason, NSF created the Innovation Corps (I-Corps) program, to encourage entrepreneurs to ask those questions as part of a “customer discovery” process. The program is based on a Stanford Univ. course called “Lean LaunchPad,” is in use at six U.S. government agencies, and has been taught globally.

I learned this from Elizabeth Good Mazhari, a certified I-Corps instructor; Entrepreneur-in-Residence at the University of Maryland; and consultant on commercializing research results. She put on a customer discovery workshop[2] for Duke (Univ.) Innovation and Entrepreneurship, in a cool space called “The Bullpen” in a refurbished tobacco factory. The name is a clever pun: Durham, N.C., made its name on tobacco in the 1800s, starting with a brand called “Bull Durham,” which was also the title of a 1988 hit movie about baseball partially filmed here (baseball pitchers sit in a “bullpen”).

Customer discovery is not “prospecting,” trying to identify potential customers to sell to them. As one of Mazhari’s slides put it, “Your job is not to validate your product… It’s to validate the problem, who has it and then how best to solve it.” She showed a short video illustrating that this is not about selling. In it, the founder answers a single question from the interviewee with a long demo of the product, and 20 minutes later has learned nothing from the stakeholder.

Elsewhere she called it a way to apply the scientific method to innovation. That means creating a hypothesis about what your customers want, and trying to prove (or better, disprove) it by talking to many strangers—not people in your circle likely to tell you what you want to hear. She says this is the real job of a startup, as illustrated by the I-Corps definition of one: “A temporary organization designed to search for a repeatable and scalable business model.” The discovery effort comes before creating your business plan or delivering a pitch or even doing much development, intended to answer these questions from her slides:

  • “Can you identify and validate a problem or need in the market that enough people care about?” (Enough to pay for, I add.)
  • “Do you have a feasible solution (innovation) for meeting this problem or need?”

Then, of course, you have to figure out if you can build a product or service whose features fill those customer needs.

The process starts with identifying all of your customer types, and here I must question why I-Corps redefines that term. What I-Corps is calling “customers,” most of the business world calls “stakeholders,” meaning people (including customers) with any interest in the product or service, in and outside of the company. I kept tripping over this word choice, so I will use the standard term. Mazhari gave the example of a machine for eye surgery to help us list the many stakeholders for the device: patients, doctors, nurses, technicians, insurance companies which pay for the service, and administrators who purchase the machine.

I-Corps would have you interview multiple examples of each of these. Do the interviews face-to-face if possible, using open-ended questions—those without pre-supplied answers like, “Yes or No.” The point is to let the stakeholders lead the conversation, rather than your assumptions leading it. Correctly done, the stakeholder does most of the talking. Mazhari had far too many excellent tips for me to capture here, but as an ex-reporter, I recognized her interviewing advice as coming straight out of my old Reporting 101 class.

Finding and talking to these people require the founders to GOTB (“Get Out of the Building”) to events where potential customers can be found. Conferences are especially fertile grounds, Mazhari said, where I-Corps companies “can knock out 20 interviews” in a few days. LinkedIn is another source she mentioned, and she urged asking each interviewee for referrals.

She said I-Corps insists that company founders be directly involved in the process. One slide quoted famed physicist Richard Feynman about scientific discovery: “The first principle is that you must not fool yourself, and you are the easiest person to fool.” It is far easier for a founder to dismiss information they don’t like from an employee than it is when heard directly from multiple customers, she said.

The objective of the exercise is to create a set of “value propositions,” one for each stakeholder type, based on the patterns of responses in the interviews. She emphasized you’re not looking for statistical validity, just patterns of responses to guide your decision-making. Propositions should be, “Quantifiable, Relevant, Specific, and Testable,” and fit one of two types, listed here with some of her examples:

  • “Pain Killers”:
    • “Reduce incidence of 30-day readmissions”
    • “25% total equipment cost savings”
  • “Gain Creators”:
    • “Provide 35% more organs for transplant”
    • “Increases 90-day survival rate from 17% to 94%”

Notice that the product isn’t even mentioned! Your focus is on benefits to the customer instead of features of the product, exactly what I found in researching my master’s thesis on persuasion in the 1990s, and preached by marketers since at least the 1970s.

If you still aren’t convinced of the need for customer discovery, I relay her cautionary tale of Juicero. This was a spectacularly public “fail,” a machine costing, originally, $700 that crushed $8 packs of raw ingredients into “fresh” juice. What the founders missed on the $120-million path to bankruptcy was that: people don’t think of something packed and shipped as fresh; Juicero’s boast that the machine featured the power to crush two Teslas meant nothing to consumers who learned they could get similar results hand-squeezing the packets; and that few people are willing to pay $8 for a glass of juice at home, on top of the machine cost.

As Mazhari of course knew, none of her information was ground-breaking. It is, however, a nice repackaging of the best practices. The problem, as always, is getting startups to actually do them. I noticed that most of the workshop participants were from one large employer instead of a bunch of startups as hoped. The full I-Corps program uses a carrot-and-stick approach, Mazhari said. All startups seeking a second round of funding from NSF are required to do a minor version of the program, and an expanded version pays for it: $50,000 strictly for market research, including a requirement to interview at least 100 prospective customers over two months. (I learned later from a founder that a significant portion of the money must be spent on I-Corps events, though.)

Sadly for me, Mazhari had no miraculous solution for motivating my clients to spend their limited funds on this kind of marketing research, even though it is guaranteed to save them time, money, and heartache in at least two ways. First, you won’t waste those on a product that seems cool but no one will pay for. Second, if there is interest, you won’t waste time developing features that add no value for customers. An approach like I-Corps may prevent your company from getting crushed like a packet of raw vegetables—or two Teslas.


Please share this post at the bottom of the page.

[1] Frequently Asked Questions (FAQs) for Small Business Innovation Research (SBIR) and Small Business Technology Transfer (STTR) Programs (Nsf19047) | NSF – National Science Foundation. Accessed 24 Oct. 2019.

[2] Mazhari, E. (2019), “Improving Innovation with Customer Discovery,” presentation at Duke Innovation and Entrepreneurship, Durham, N.C., 10/3.

Tell the world: