I served as interim CTO for a startup developing a consumer product for families. The founder was analytically minded, but several product decisions rested on his personal beliefs about how households make decisions and which features would motivate different users.
The development team could implement those ideas, but discussion could not establish whether the underlying assumptions were correct.
The Problem
Some disagreements concerned familiar interface conventions. More important were the founder's assumptions about customer behavior. We were at risk of building a consumer product around one person's household rather than the market.
Repeated argument did not help. The founder did not find my interpretation persuasive, and neither of us had evidence.
The Experiment
I proposed a modest advertising test before committing more development money. For several demographic groups, we created four messages emphasizing different potential benefits of the product.
The application was not yet available for purchase, so the experiment could not measure sales. It could measure which messages attracted clicks and what visitors did on the corresponding landing pages, including engagement and time spent.
This gave us behavioral evidence about the propositions that resonated with each group. The goal was not to optimize advertising; it was to test the product assumptions embedded in the roadmap.
The Result
The results showed that some proposed benefits attracted the intended users and others did not. Questions that had been impossible to resolve through discussion could now be evaluated against observed behavior.
The experiment cost little compared with implementing the disputed features. It gave the team a basis for prioritization and reduced the risk of building an expensive product around an untested conviction.
For an early consumer product, evidence does not always require a finished application or a large research program. A small, well-designed market test can be enough to replace "I just know" with something measurable.
See other Early-Stage Focus: One Business Model
