How you choose who to survey matters as much as how many people you survey. A few common approaches:
Every member of your population has an equal chance of being invited. The gold standard for representativeness, but requires a full list of your population to draw from.
Surveying whoever is easiest to reach (a public link on your website, for example). Fast and cheap, but the people who see and respond to it may not represent your whole audience - be cautious generalizing convenience-sample results to a broader population.
Splitting your population into segments (region, plan tier, tenure) and sampling proportionally from each, so smaller but important segments aren't drowned out by your largest group.
For small, known populations - your whole team, all customers of a niche product - it's often easier and more accurate to just survey everyone rather than calculate a sample.
Whichever method you choose, use the Sample Size Calculator once you know your population and desired confidence level, and set expectations early: a convenience sample with a great response rate can still beat a "properly" random sample that mostly goes unanswered.