Analytics
1.5K members online now
Understand information in your reports and troubleshoot reporting issues such as self-referrals, (not set) data, and inaccurate information
 
Guide Me
star_border
Reply

Data Sampling

N/A
# 1
N/A

I'm receiving the message "This report is based on 63.65% of sessions.".  However, I only have one view in my property, and there are only around 60k sessions for the selected date.  I thought sampling only occurred after 500k sessions?

 

Thanks,

Ryan

Data Sampling

Explorer ✭ ✭ ☆
# 2
Explorer ✭ ✭ ☆

Are you sure sampling is set to "Greater precision"? Also, last time I checked, there was a bug in the GA interface where the very 1st report was always using "Faster response" even if it said "Greater precision". Refresh the report or install Da Vinci to fix that.

 

Can you provide more details about your request. Any segment? filters? which report?

Re: Data Sampling

N/A
# 3
N/A
Thanks for the reply. Yes, greater precision is set. I get sampling in both the website and API.

The property has a single view with user id enabled. There are no filters. I'm using a tracking pixel in our email marketing to record opens via event and campaign parameters. In addition, I have a link tracking script that uses the measurement protocol to record clicks.

So basically, in the report I'm trying to look at event action, category, and label values, along with campaign parameters to get all opens and also all clicks.

The default total event reports are not sampled. It uses sampled data when I add a secondary dimension, create a custom report, or use the API in an attempt to retrieve all of the detail rows. I do not believe there are ever more than around 60k rows for a given day.

Thanks

Data Sampling

Explorer ✭ ✭ ☆
# 4
Explorer ✭ ✭ ☆

Hi there,

Keep in mind sampling is impacted by the total number of sessions, not only those with events. If you add a campaign parameter (or anything else) as the secondary dimension, it can easily lead to sampling. Earlier you mentioned "60k sessions" - but in your latest comment you say "60k rows", thus 60k different event category/action/label (aka "cardinality" of the table). Either way, on its own, there shouldn't be sampling. But if you add a second dimension, you might.

Re: Data Sampling

N/A
# 5
N/A
Hi Stéphane,

Yeah, if I look at sessions for that day on the Analytics home page, it reads 60K. That being the case, and the fact that there shouldn't be any sampling, is there any way to get this corrected?

Thanks,
Ryan

Data Sampling

Participant ✭ ✭ ☆
# 6
Participant ✭ ✭ ☆

I constantly have this problem in Google Analytics. I was using the Google Analytics plug-in to Google Sheets for a while, though it was kind of clunky (though free), and you can shorten your dates ranges to see non-sampled data. What I do now is I've subscribed to SuperMetrics for about $50/month. It's not perfect, but it's a lot easier than the GA plugin and the data pulls can be automated. It can take a little work to stream together the data so that it's not sampled, but it's better than just trusting GA's guess (in my opinion).

Data Sampling

Explorer ✭ ✭ ☆
# 7
Explorer ✭ ✭ ☆

Great - there are a couple of tools to retrieve data. The one I've been using for years is Analytics Canvas - the benefit is it's closer to a real ETL tool (Extract-Transform-Load).

 

Disclaimer: I have no affiliation to Analytics Canvas beside being a happy customer Smiley Wink

Data Sampling

N/A
# 8
N/A

Thanks for the recommendations.  Unfortunately, I get sampled data even if shorten my date range down to a single day.  Do either of those products get around that in any way?

 

I'm getting desperate, as my deadline for this project is rapidly approaching.  I'm afraid I'm going to have to look at other alternatives instead of Google Analytics if I can't get around this issue.

 

I appreciate the help.

Data Sampling

Explorer ✭ ✭ ☆
# 9
Explorer ✭ ✭ ☆

The smallest granularity available is by day - although in theory you could build you own extract by hour.

If it's a very large account, consider using Google 360.

 

But before that... if you want, let's go back to the root cause: What is the business question you are trying to answer? Maybe there's a totally different way to approach the problem. If you need further help, you can also hire help specifically to tackle this challenge - doesn't have to be me... I'm already busy enough Smiley Wink

Re: Data Sampling

N/A
# 10
N/A

Here's what I'm trying to do: record marketing email newsletter opens and clicks by utilizing event tracking and campaign parameters. Then, I want to be able to pull reports that list each newsletter open and click with accompanying information like the name of the newsletter, the subject line, distribution date, open or click time stamp, link URL, etc.

 

I understand you are busy, so I appreciate the time you've spent with me.  I wish I could go straight to Google for help, but it's hard to complain about a free product.

 

Most specifically, I'd like to understand why my data is being sampled.  From my understanding, it should not be as there are well below 500k sessions for the period.  I'm either misunderstanding something, there is a different reason sampling is occurring, or there is a problem with my account.  Here is a screenshot:

 

Capture.PNG

It's saying the report is based on 63.5% of sessions, but it's only reporting 60,203 sessions.  Am I misinterpreting something?

 

You mentioned building an extract by hour.  How can I do that from the API when it only accepts yyyy-MM-dd as the to and from dates?

 

Thanks again.