Re:Ask HN: How many responses should I wait for when testing...
by 2manyredirects on Thursday April 20, 2017 @12:49PM
(#14158364)
That's super-awesome advice @brudgers, thank you.
I was so fascinated with the survey results being statistically significant (and therefore thinking the more respondents the better) that I failed to realise that 5 people willing to take the time to respond, and 4 willing to share their emails is such a good starting point to get the ball rolling.
Five people is an incredible amount of interest: The median amount of interest in a project is zero. Take those five people seriously. Treat them like the most important people that your business has ever come across...because they are. The actually might care.
Responding to those five people's input is how a business develops a culture of listening to its users/customers/potential customers. Ignoring their input is how it develops a different habit.
Five people might be manageable for a individual's side project. Five hundred people won't be. Make something those five people want. It is good practice at making things people want. Making tweets is easier than that. Don't make more tweets until you have something worth tweeting about...worth tweeting about in the sense that people will care much more about it than what got five people to respond.
Facebook started on a hall in a dormitory. It met an unmet need of just those people.
Good luck.