All Collections
Understanding SEO Performance
Non-brand Organic Traffic
How can Google Analytics data sampling affect your campaign?
How can Google Analytics data sampling affect your campaign?

The traffic data from your Google Analytics can be different from the one shown in SEOmonitor's Organic Traffic section

Updated over a week ago

If you're looking at a website with more than 10 million sessions (referred to as "hits" in their T&C's) per month, Google Analytics (GA) would apply session sampling to your data (unless you are paying for their Premium version). This is a way of selecting a subset of the actual data in order to save time on processing all of the data.

There is a warning shown in the GA interface and you'd find the corresponding one in SEOmonitor's Organic Traffic module as well:

The data sampling is justified on their side as a way of providing your reports in a timely fashion and it refers to both sessions and conversion data.

How can you check?

Misalignments between GA and SEOmonitor would occur when checking timeframes, rather than daily values.

Even in GA alone, there would be a difference between the information shown for a certain day, when hovering over a longer timeframe selection, and the one shown per day, when it is only that exact day that is selected. So, the larger the volume of data, the more the chances of sampling.

Both the longer timeframe selection and the individual day selection would "suffer" from sampling, but the day-by-day one would be less sampled. The data we extract from GA is at a daily level, so even if sampled, it is as comprehensive as possible.

You'd need to make sure that in GA you have selected:

  • Google as the Source;

  • Organic Traffic as the Medium;

  • the exact same date in the calendar.

The usual pattern of behaviour would be to check the reports for longer timeframes, for example one month. For the reasons mentioned above, the monthly totals from GA would be different from the ones in SEOmonitor for these longer timeframes, so a better comparison would be to cross-check for each day separately.

Did this answer your question?