Etsy has a small article about the difference between Shop Stats and Google Analytics. It’s a nice insight into how Shop Stats works but it includes one very interesting statement:

Google Analytics samples data, which means they record a subset of the data and use statistical analysis to estimate the actual numbers.

Well that doesn’t sound very good! But is it true?


Sampling is indeed a big problem for large organisations that use Google Analytics.

For custom reports, it happens when the date range being analysed contains 250,000 or more sessions (regardless of whether you need them all in your report).

In the standard reports (where you spend most of your time), it happens when a report has more than 50,000 rows. The extra rows are aggregated into one called “other”. These are rows in a report (normally dimensions, not metrics):

Rows in GA reports are normally dimensions like source/medium or landing pages.

Does it apply to me?

So 250,000 sessions… 50,000 rows… those are some big numbers.

To give you some context, you’d need to be getting at least 20,000 visits per month to hit this limit on a report displaying 12 months of data. That’d probably be something like 18,000 people every month, looking at your shop.

If you have those kinds of numbers, we need to talk.

But even if you did have those numbers, that’s a report over 12 months. We don’t analyse that time period every day.

For standard reports, it’s based on the number of unique values within each dimension (that’s the 50,000 limit), which are aggregated on a daily basis.

So in one day, you’d need to get:

  • Traffic from 50,000 different sources
  • Views to over 50,000 different URLs in your store
  • Visits from over 50,000 different locations

I don’t think your data is getting sampled. Let me know in the comments if you think this is happening to you!

Comments (1)

Thank you! I saw this on Etsy the other day and was like “Excusi?” Glad you gave the whole story. If only I had those numbers!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.