Our real-world testing was highlighted in the Harvard Business Review
Validate-first, build-later requires product innovators and marketers to get their hands dirty with testing when the product is still just a tiny spark of innovation.
True or False: you have to build a product to know if someone wants to buy it.
True for Henry Ford. He built his first car, the Quadricycle, in 1896 while working as an engineer for the Edison Illuminating Company and used the proceeds from selling it to build his next car. Market research didn’t come along till the 1920s, and it’s not clear that it would have done much for Henry anyway: truly breakthrough products need to be experienced for consumers to respond, not just described in a focus group or survey.
Roll forward 125 years. Build-first, validate-later has echoes in Tesla and other capital-heavy game-changers, but it’s no longer the rule for most products. Thanks to the magic of great design, most products can be rendered in a way that allows consumers to make judgments about them when encountered online, still in the “coming soon” stage.
So we’re going to go with False.
Validate-first, build-later requires product innovators and marketers to get their hands dirty with testing when the product is still just a tiny spark of innovation. They need to agree on product definition, document possible value propositions, riff on different concepts for fulfilling the product vision, and then package it in a way to attract an authentic and measurable reaction from customers.
It’s messy and hard. But it has all kinds of great benefits.
Here are five reasons to test before you build:
It’s almost always cited as the primary reason new products fail: no market demand.
How do new products get launched without validating demand? In no particular order:
- taking focus groups too seriously
- prioritizing speed over all else
- assuming surveys indicate purchase intent
- confusing market size with product demand
- hubris
to name a few.
Validate demand for your product as soon as you can—and in most cases, that’s before you build the product, not after. People who click on ads for your product-to-be are a better indicator of interest than people who take surveys. And people who give you an email address to be notified when you launch are pure gold.
Bye-bye Lean Startup?
The Lean Startup mantra: build a minimum viable product, get people to try it, rebuild it if it doesn't work, and try again. The Lean Startup sparked a revolution in the way people thought about product development. You can fail and start over! Launching is learning! You don’t have to figure it all out up front!
It’s all still true. You just don’t have to be an engineer to make it work.
Testing new products while they are in the development stage provides feedback on customer need, product features, and value proposition. But you don’t need to build three versions of a product to test it: you can test visualizations of a coming-soon product with real customers (be honest about the coming-soon part) and measure their relative interest in different versions through clicks on ads or email signups.
Testing: a cheaper way to de-risk the product development process.
(Bonus reading: this thoughtful analysis of how the Lean Startup model almost killed a company.)
Companies are brimming with good ideas for new products. And they are often paralyzed when it comes to prioritizing them. The result: stasis.
Side-by-side testing of new product concepts with a range of target audiences provides data about which ones to start with, which ones to shelve and revisit, and which ones to abandon.
What kind of data? Data like click-through rates and landing page views about relative interest in a new product offering. Data about the proportion of people in a target audience who sign up to learn more or to be notified about launch. Data about new responsive audiences that expand a brand’s footprint.
Resources are limited. Data generated via testing helps allocate them wisely.
Where new product ideas go to die: corporate politics.
So many great new product ideas are squelched by opinion, not data. Change up your culture by opening up new ideas to testing. If a product concept is a dud, you will know with certainty; if it is in fact redemptive, you will be a hero. Either way, the decision will be based on facts, not power plays.
Were you responsible for Cheeto’s Lip Balm? We’ll keep your secret, but we hope you’ve changed your ways.
We’re all for failing fast, but why fail spectacularly? Test first, people. Make sure a high proportion of your target audience wants your product so badly that they are willing to share an email address with you.
Also, maybe Cheeto's Lip Balm would be a huge hit today. Have you seen the Chipotle makeup collection that just debuted?
