Prototyping can increase a startup's chances of success from 5% to 80%.
But first, a quick detour:
I spent the weekend in Sonoma and Napa with my wife, celebrating our first anniversary. Deb and I enjoyed two days of elegant dinners, beach time, and some glou-glou in the great al fresco.
So I'm relaxed AF as I write this. Here are the obligatory photographs.
I've been married for only 370 days, but that won't stop me from opining: the key to a successful marriage isn't communication, or trust, or gratitude.
Prototyping is a small effort that gives you a signal.
More specifically, a prototype is a small facade that represents a real thing, which you put in semi-real conditions with real customers to gather signals that you're on the right track.
When it comes to marriage, there are a long string of prototypes leading up to it: chatting, dating, weekend trips, closet space, cohabitating, and engagement.
In fact, relationships are just a chain of prototypes, where you slowly increase the size of the bet in proportion to the evidence.
I know: I'm a romantic.
And each prototype gives you a signal about whether you're heading in the right direction, so you can pivot (or bail) when the balance of evidence tips the other way.
And that's why we prototype:
To get rapid feedback from the people we need to hear it from, with as little effort as we can.
It's about starting with small bets, and slowly increasing the size of each bet as we get more data --- and more confidence.
Here's an amazing stat: 95% of startups fail. But with just 10 prototypes, that failure rate drops by almost half.
By the time you reach 50 prototypes, the failure rate is under 20%.
That's the power of prototyping.
It's why prototyping is at the very core of my Traction Thinking philosophy, and why I bring it into my community work, into my coaching, and into everything we do at The Right Box.
And, despite popular belief:
You can prototype anything.
- and more.
Prototyping is not about creating a small version of the whole thing.
It's not an MVP.
It's about a creating a version of the thing that allows you test the "surface" of customer interaction. Prototypes are a facade.
In a western film, when the stranger walks through the town, his spurs clanking on his boots behind him, the buildings beside him aren't real --- they're facades.
But they're enough for us to experience the town without building it.
Let's talk examples.
How do you prototype a value proposition (i.e. an offering)?
- Customer interviews
- Landing page
- Social media ads
- Paper sketches
- Google Slides
- Clickable prototypes (e.g. in Figma)
- Zapier-based Frankenstein's monsters
- Paper sketches
- Individual touchpoints (e.g. emails)
- Manual, concierge delivery
You get the idea.
The goal in each of these is to test the surface of interaction --- the part the customer sees.
But you don't prototype the whole thing.
You are looking for the moment of impact --- the part delivers the most value to the customer. The part that has the most risk. The moment that matters.
Back to marriage:
Conflict is inevitable. There are many kinds of tools and techniques couples can use to help resolve them. But just busting those out for the first time in the middle of a tense fight doesn't set you up for success.
The solution? Prototypes!
Use light versions of the tools during good times. Then use them during smaller disagreements. Slowly progress toward the larger conflicts.
But only if confirmatory data comes in. Otherwise, pivot to other tools.
In other words, you don't have to sit around and wait for the right circumstances (like a big fight, or a launched product) to put the idea out there and get a signal back to show if you're heading in the right direction.
In your head right now, you may be thinking, "well, you can't prototype this-thing-in-my-mind".
Yes, you can.
Here's another example. The startup Savioke wanted to build a robot with personality that would deliver concierge items for hotel guests, instead of requiring on hospitality staff.
They wanted to see how customers would react to the robot...which didn't exist.
Savioke had an existing robot, though it did a fraction of what it would need to. So they decided to control the robot remotely using a PlayStation controller, and strap an iPad Mini to it, which would show a series of slides to mimic an interface.
Give it a soundtrack and fake phone call with guests, and... 💥 Prototype.
They did this in less than one day.
You really can prototype anything.
Still don't believe me? Reply to this email with your problem and I'll tell you how to prototype it.
So! Your mission, should you choose to accept it is to launch a prototype within the next week and put it in front of real customers.
There's no reason not to do it. And every reason to do it.
Bonus! 3 prototyping pro tips
Here are three things you can do massively level-up any prototyping effort.
Pro Tip 1: Proportion your effort to the evidence.
Prototypes are a spectrum, from a screen sketched on paper to a moving robot.
How do you pick where on the spectrum to prototype?
Go where the evidence is.
- Little evidence you're on the right track? Only expend a little effort.
- Tons of evidence? Maybe you can afford more effort.
Only expend as much effort as you have evidence.
Pro Tip 2: Know what you know --- and what you don't.
In order to accurately assess how much evidence you have, you need to understand what is an hypothesis (or assumption) and what is a fact.
Rank what you know on a scale of how much you know it, from wild-ass guess to reasonable inference, to a little data, to robust data.
Pro Tip 3: Look for a big improvement to the status quo.
The biggest mistake you can make is building for "meh".
When you put your prototype in front of customers, you're looking for a big wow. You want their eyes to light up.
Just a little bit better than the status quo doesn't cut it.
Pivot and keep trying until you get that "wow" reaction.
By the way...
I'm working on some fun new resources to help with prototyping. If you want to dive deeper, click here and you'll be the first to know when they launch.