Data Studio
5.2K members online now
5.2K members online now
Ask questions about Data Studio
Guide Me
star_border
Reply

Scorecard reports different numbers than time series in DataStudio

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Hello,

I am trying to understand what causes my dashboard to show different numbers for the same metric if I put it on a scorecard (set for "today") and a timeseries chart (that includes today's date). 

In the below example I am showing the (supposedly) same number in 3 different ways - scorecard set for today, timeseries set for today and timeseries set for "last 7 days including today". All 3 show different results.

 

Would appreciate to get some help on understanding the discrepancies

 

*Note - the example is based on a metric specific to our product (passengers check-ins)

comparison.PNG

 

1 Expert replyverified_user

Scorecard reports different numbers than time series in DataStudio

Google Employee
# 2
Google Employee

Hi Nadia,

 

This sounds similar to the questions asked in this thread:

 

https://www.en.advertisercommunity.com/t5/Data-Studio/Scorecard-reports-different-numbers-than-time-...

 

Please take a look and let us know if that answers your question.

 

Best

 

Rick

Scorecard reports different numbers than time series in DataStudio

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

Hi Rick,

Thank you for a quick response. This is indeed a similar question, but also significantly different. The previous question was about users, and the explanation was lying in the difference between users and unique users, whereas in my example we're looking at passengers check-ins, and there is not element of repetition here.

In anyway I parse my numbers I expect to see the exact same number of passengers checked in for a specific day. That's what's so weird - looking at passengers checked-in in 3 different ways I get 3 very different numbers. How is that possible?

* Is there a different in the way a scorecard defines as day (time period) and a timeseries defines a day?

* If yes, then how is it possible that I get different numbers for a specific day through timeseries in which this day is included and a timeseries that shows only that day?

Scorecard reports different numbers than time series in DataStudio

Google Employee
# 4
Google Employee

Hi Nadia,

 

Scorecards and time series should not be defining time periods differently, though as noted in the referenced thread, scorecards aggregate the metric across the time frame, whereas time series will sub total for each day for each unique instance of the data.

 

A couple of questions to consider:

1) Do either the scorecard or times series have filters set?

2) Do either the scorecard or times series have default date ranges set?

3) Are Do either the scorecard or times series grouped with another element on the page?

4) Is there a breakdown dimension set for the time series? 

 

Any of those things might cause data to display differently.

 

Let me know if any of those lead to an answer!

 

Best

 

Rick

Scorecard reports different numbers than time series in DataStudio

[ Edited ]
Visitor ✭ ✭ ✭
# 5
Visitor ✭ ✭ ✭

Hi Rick,

Thank you for the thoughts and suggestions. Please find an answer for each here below. Also I've attached a picture of the dashboards in question along with a screenshot of their properties panel, for better clarity.

reports and properties.png

1. Neither the scorecard nor the timeseries are filtered - they both are very simply defined with dimension="Date" + metric="Passengers Checked in".

2. The scorecard is set to "Auto" time range (which is tied to the page Date Range controller - set to "Last 14 days (include today)"). Whereas the scorecard is set to custom time range (defined as "Today")

3. Not sure about the "grouped" question? Could you please clarify what that would mean?

4. No breakdown dimensions for the timeseries.

Scorecard reports different numbers than time series in DataStudio

Google Employee
# 6
Google Employee

Hi Nadia,

 

I looked into this a little more. I think the issue is that data for "Today" is generally not complete until it's no longer today, if that makes sense. Network latency, backend aggregation jobs completing at different times and similar arcane things can mean that two views of the data won't always agree until all those things have resolved.

 

What is the data source for this report?

 

Best

 

Rick

Scorecard reports different numbers than time series in DataStudio

[ Edited ]
Visitor ✭ ✭ ✭
# 7
Visitor ✭ ✭ ✭

Hi Rick,

Sorry for the belated reply, it took me a while to get to putting the whole picture together and doing some follow-up on the numbers, examining your suggestion above.

Unfortunately I can't see how your suggestion would settle the discrepancies. Even when the day is coming closer to an end, the numbers do not come closer together.

 

To your question - our data source is GoogleAnalytics - the "bucket" collecting ata from our mobile app.

 

Here's what I could see when did more investigation of those discrepancies:

Seems like the discrepancies show only when working with custom metrics (for example - "Passengers Checked-in" in my investigation and screenshots).
It seems that the custom metrics results change based on the time frame (7 days, vs. 14 day, vs. 28 days etc.). Therefore when we're comparing scorecard result to the timeseries result, we're getting different numbers for the day to which the scorecard is set. If the timeseries is set to the same timeframe / day of the scorecard, we'll get same numbers.

When comparing the numbers in DataStudio vs. the numbers in GoogleAnalytics - the data source (using dashborad, since they're based on 100% activity vs. the custom reports that are only based on partial sample) - we find same discrepancies here, and again, only for custom metrics. Once the GoogleAnalytics dashboard and the DataStudio dashboard are set to the same timeframe same number will show, otherwise, the numbers would not match.

* Please see more specific examples and screenshots below.

 

It does really look like some sort of bug, or system inaccuracy with the way DataStudio parses numbers within a time range.

We are trying to use DataStudio more and make it our main tool to measure application health and performance, however, we need to know that the numbers are reliable. It would be really helpful if you could look into this further and advise what's wrong, and possibly suggest a remedy.

 

Sessions: - this the example for Google Analytics standasrd metric (non-custom) showing that all numbers are consistent for non-custom metrics
DataStudio:

screenshot-2.pngDataStudio - Passengers Checked-in
Google Analytics:

screenshot-3.png


Passengers check-in: - inconsistencies for the custom metric numbers, in accordance to the chosen time-range
DataStudio:

screenshot-4.pngDataStudion - Passengers Checked-in 
Google Analytics:

screenshot-5.pngGoogleAnalytics - Passengers Checked-in
DataStudio same visualization, different time-ranges:

 

screenshot-6.pngDataStudion - Passengers Checked-in - various time ranges