This post originally appears on our Medium page. Please give us a follow!

When I started a smart mailbox software company in 1998, the first thing we did was a survey. We intercepted families standing in line to see DC sights, and asked about their willingness to put a smart mailbox in front of their homes to receive e-commerce packages.

Guess what? Our solution was a big hit. A no-brainer. Or so they said. Nearly 20 years later, there are still zero Internet-connected mailboxes receiving and sending e-commerce packages for single-family homes. And to this day, I’m no fan of surveys, especially early in the life of a new concept.

“Progress” for a new venture means validating the untested assumptions. As we design and execute our in-market validation experiments, innovators must be careful not to take false comfort in “Say” data (versus “Do” data).

Here’s a framework that compares some of the options available to test your assumptions:

The arrow shows the entrepreneur’s preferred path from hunch to launch. Experienced entrepreneurs have a strong bias to stay in the domain of what users Do, rather than Say.

Of course, as an intrapreneur, you work within the risk-averse world of a core business. You must adhere to your organization’s decision-making process, which usually requires quantitative validation. That’s OK, we welcome input from the Say domain, as long as it used to augment our Do data, not to replace it.

And when the two are in conflict? To us, Do beats Say every time.

But that doesn’t mean it’s an easy conversation. Have you seen conflict between data on what users Do, versus what they Say? How did your team navigate that conflict? I’d love to hear your insights.

Post a comment here, or reach me at