Google Analytics Sampling[ Edited ]
May 2017 - last edited May 2017
When using the _setSampleRate, does GA automatically adjust reports via the GA dashboard interface to calculate the results?
For example, when looking at the page view count with a _setSampleRate at 50%, would this be the total number of page views from the sampled selection, or would it show the views*2 for the expected total page views a website has?
Google Analytics Sampling
That's a great question. Adjusting the sample rate affects all your standard reports - all your data is sampled, even if it does not say so in the interface.
I've never personally used the sample rate modification but I'd assume that the impact is the same as looking at a sampled report. i.e. your page view report, with a 50% sampling rate, will be created by looking at 50% of your data; with all the unreliability that entails.
- _setSampleRate is legacy code, you should plan to migrate to universal. I'd recommend doing this via Google Tag Manager.
- adjusting sample rate is pretty extreme as your are forever manipulating your data with no fall back to benchmark the impact of sampling. Do you need to do this? If the amount of data you are sending to Google is too high (i.e. greater than 10 million hits/month) then consider removing, for example, excess hits which could come from events or virtual page views. If your data collection comes from the same GA code on several different sites then consider implementing separate properties rather than one. If the amount of sessions is the issue then you could look at making a case, based on the revenue your site generates or its importance to the organisation, for Premium (150 000 USD/year) where you do not have to worry (so much) about how many hits you send.
Was my response helpful? If yes, please mark it as the ‘Best Answer.’ Learn how here.
Google Analytics Sampling
Thanks @John W for your response. This is super helpful. I understand that all reports and data will become sampled reports and less accurate as a source of analysis, however it's a short term solution right now as I'm looking at over 10 mil hits per month.
Would the reports in the dashboard show lower numbers when using sampled data (not taking into account any sampling error). I'm wondering if changing from a premium account (sending all my data) to a free account (changing it to 50% of my data) will show a 50% drop in page views, or if Google accounts for the sampling rate and will adjust accordingly. I understand percents won't be impacted (minus being more inaccurate), but primarily concerned about the calculated "raw" numbers, ie. 654,321 page view.