A 300-person survey that beat a 5,000-person one

A 300-person survey that beat a 5,000-person one

A niche dental-software company couldn't afford a giant panel, and their last big-sample survey had flopped. The setup was the assumption that bigger samples win press. The angle was the opposite: survey a narrow, hard-to-reach group, in this case 312 practicing dentists, and own a question nobody else could ask.

The pitch leaned on access, not scale: 'We asked 300 dentists what procedure they secretly think is overprescribed. 58% named one, and it's not the one you'd guess.'

The result: 29 placements including a Vox health piece and an NPR affiliate, 23 with links, all because the respondents were genuinely hard to assemble. The survey page pulled 7,800 sessions. A health reporter noted, 'Three hundred real dentists is more interesting than five thousand random adults guessing about teeth.'

We disclosed the sample size openly in the pitch, which built trust instead of inviting suspicion.

Lesson: Sample size impresses statisticians, not editors. A small, exclusive audience asked the right question beats a giant panel asked an obvious one.