Skip to main content
Back to articles
The Voxaa blog

Remote concept tests: what actually works

MethodJune 20267 min read

Testing a concept remotely works, provided you hold three constraints: one cell per respondent, a stimulus that reads on a phone screen, and a sample sized for the difference you are trying to detect.

Monadic by default

Showing several concepts to the same respondent creates a comparison that does not exist on the shelf. The monadic design, one concept per person, costs more in sample but measures what really happens at the moment of choice. Sequential monadic is an acceptable compromise on two concepts, never on four.

The stimulus has to survive the phone

A board designed for a test room becomes unreadable on a six-inch screen. One visual per message, text above sixteen pixels, no meaningful detail in the margins, and a video under thirty seconds: beyond that, drop-out climbs before the first question.

Size for the expected difference

One hundred and fifty respondents per cell can detect a ten-point difference in purchase intent. Below one hundred, the confidence interval covers most differences between concepts, and the test concludes there is no difference where it simply lacks the power to see one.

Key takeaways

  • One concept per respondent: monadic remains the reference.
  • Validate the stimulus on a phone, not on a desktop screen.
  • One hundred and fifty respondents per cell, or the test concludes nothing.

Read next

Get our next articles

One email a month: methods, figures and field notes, no advertising.