1,000 survey respondents. Still couldn't answer the question.
Here's why.
The analysis plan called for breakouts across 6 cities, 3 age groups, and 2 usage segments.
Divided that finely, some cells ended up with 25 respondents.
You can't make a statistically defensible claim from 25 respondents. You can't run significance tests. You can't compare segments. The sample was large in the wrong places — and couldn't answer the question it was designed to answer.
The counterintuitive fact most people don't know:
Once your population exceeds roughly 50,000 people, the required sample size barely changes.
A survey of a city of 200,000 needs almost the same number of respondents as one representing 1.4 billion — at the same confidence level and margin of error.
At 95% confidence, ±5% margin of error: 385 respondents.
Not 3,850. Not 38,500.
The question that should come before any sample calculation:
What decision does this research need to support — and what is the cost of being wrong?
Sample size is a function of decision risk. Not a round number. Not a budget default.
→ Full guide: maction.com/survey-sample-size-calculator