How to build a prototype: Choose just one thing

Sep 25 2014

As our student startup teams embark on building a first prototype of a product idea, a common issue emerges: prioritizing what should go into the prototype. Something must go.

Actually, almost everything should go.

So in the Lab, we ask students to answer this question: What’s the one thing you need to know right now?

I feel like Curly from the movie City Slickers as he tries to explain the secret of life. Holding up one finger, Curly tells Mitch, “One thing. Just one thing. You stick to that and the rest don’t mean $!%&.” We help students identify the one assumption they have that if incorrect, would make their product pointless. That changes everything.

For example, one group of students is working this semester on a project involving restaurant sanitation grades and food allergies. The students’ first inclination for the prototype was an interface that displayed the pertinent information about a restaurant.

But that misses a big assumption. Would anyone even want to get that page? Is there sufficient perceived value for customers to take the first step and enter a restaurant’s name for information?

If the customer will not take that first step, how (or even if) the data is displayed really doesn’t matter. It’s the one thing: Will customers provide the name of a restaurant they want to visit in order to get more information?

To find out, students must conduct a simple test of a specific action to determine if customers want what’s being offered.

You don’t need a fully functioning database to determine if customers want to search through it. You only need the input fields for an email address and search terms. Will customers feel tricked if you give them a “sorry, we really don’t have the database” page? Possibly, yes. But you should be willing to find that information by hand and get it back to them as quickly as possible.

By doing so, you’ve proven there is enough value for the customer to take the first action and that she will give you something of value (albeit limited): an email address. When you deliver the results of your manual search, you have another opportunity to engage with the customer to learn more about what she needs and expects from the service.

The same holds true for a group of students working on a Teach For America model for journalists. Their first assumption was that students would sign up because they’re guaranteed a job. It’s a great assumption, but it must be proven. If students aren’t interested, the next step doesn’t matter.

What about media organizations in which the students would be placed? Our assumption is that the organizations would take the students. That could be the one thing to test. As Curly told Mitch, “that’s what you have to find out.” Our students must come up with their “one thing” to test. If two things hold equal importance (as with this group), they must pick only one thing to test at a time.

“One thing. Just one thing. You stick to that and the rest don’t mean $!%&.” When that’s done, you define the next “one thing.”


Leave a Reply

Your email address will not be published. Required fields are marked *