Surveys are running on a weekly basis. Data is weighted and updated in the dashboard on a daily basis.

First month

We aim for a minimum of 1000 respondents who are active category users in the first month (if your incidence rate is >40%).

Every month following

We aim for a minimum of 340 respondents every month following. Depending on the category penetration, we may even sample more than 340 (don't worry — there's no extra cost) 🎉

We combine this into a 3 month-roll to make data points in the dashboard more robust. In the US, the 3 month-roll will equate to roughly 1,000 people.

We have increased our standard sample in the US not to decrease the margin of error but to allow our customers to have larger samples in the regions to look for regional differences with confidence.

Wondering how we know this sample size is big enough?

At the kind of population sizes we’re working with, the size of the population has no impact on the accuracy of the estimates from a sample. The only thing that impacts accuracy is the size of the sample itself. i.e. a 1,000 sample gives us a 3.1% margin of error no matter what country we are surveying in.

Increasing the sample size will improve accuracy but only marginally at sample sizes above 1,000. Increasing the sample size by 50% to 1,500, for example, only reduces the margin of error to 2.5%.

What's the margin of error on the survey?

What geography are you interested in?