Does the Google Analytics connector bring in full data or sampled data?

[asmith]

This might seem like an odd question but does anyone know if the Google Analytics connector brings in sampled data or complete data? On large websites, Google will sample data to lessen data pull loads. However, this isn't helpful if you need accurate data (eg. sales, quantity, etc.). 

Thank you!

[curth · Accepted solution]

We request all of the data. If the underlying data is itself sampled or incomplete, then of course we'd have no control over that -- but we will get everything that the Google API can give us.

[adelheni]

Hi,

So Power bi's connector for google analytics has a sampling issue and an API Calls limit. The standard limitations and quotas for Google Analytics API requests are documented in Limits and Quotas on API Requests. However, Power BI Desktop and Power BI service allow you to use the following enhanced number of queries. Queries per day: 1,500,000 Queries per 100 seconds: 4,000 Might be better to connect Google analytics to a data warehouse such as Google BigQuery, which will help you get access to historical data that you won't be able to get with Power bi's connector. I'd suggest using a third-party connector.

[soham2000]

Hello,

I am using the GA connector for power bi.

Then I am exporting the data from powerbi in excel.

Exporting GA side data into excel from GA.

Then mapping both these excel files using VLOOKUP for transaction or order ids.

I see a lot of orders which are there in the GA side are missing in the powerbi side.

So, why don't we get the full transaction ids and the full set of revenues from the GA to the power bi side.

Also, we are using power BI pro account.

So, can someone kindly help? Is the issue due to sampling? Or how to resolve it?

Thanks in advance.

[intrasight]

Because of the way this connector works, if you wish to avoid sampling, you'll have to go elsewhere. See

https://community.powerbi.com/t5/Desktop/Google-Analytics-connector-missing-startdate-enddate-parame...

[curth · Accepted solution]

We request all of the data. If the underlying data is itself sampled or incomplete, then of course we'd have no control over that -- but we will get everything that the Google API can give us.

[jellegaa]

I would suggest to use https://scitylana.com instead. This will pull all raw data unsampled from GA (free edition) into Power BI.

[Luke]

Is there any indicator within Power BI Desktop that denotes that the data is sampled?  The API would at least tell you it's not bringing in the full dataset, I believe.

[Seth_C_Bauer]

No indicator. Interestingly enough this thread has an admin answer of the connection (content pack) only pulling the last 6 months. Here

[Luke]

Thanks for the reply and related link - it's great to know the Desktop connection is the same as the one used in the content pack.  The content pack doesn't fit our needs because of the year-over-year analysis needed, but I still find it strange that no indicator is provided for sampled data in the Desktop app.  Hoping to see this connection improved in a future release.

[intrasight]

They can only get unsampled data if they use the unsampled API - which is only available to GA Premium customers.  It would be nice if PowerBI would add a new metric to indicate if sampling is present. That's what I did for my custom query integration.

[Luke]

To dig into this more, are you saying that PowerBI's current GA connector is configured through an API that may sample, despite the underlying profiles being GA Premium accounts?  With your comment on custom query integration, do you have a workaround to take advantage of the GA Premium account (and unsampled data) that does not use the PowerBI-provided API?  I'm not terribly familar with GA's API suite so any clarification would be much appreciated.

[intrasight]

Sorry for the very late response. I wish the "email me when someone replies" was checked by default.

Sampling occurs when using the GA API (which is what Power BI does). But that sampling behaves differently for Premium accounts. Only using the GA unsampled API can you avoid sampling.

Sampling mostly manifests itself when you have a wide timeframe. In the GA API you can specify the timeframe explicitly. In my custom connector, I split a query into subqueries with smaller timeframes, which can avoid many sampling scenarios. I'm actually about to post a new discuss message inquiring about how to specify the timeframe using the PowerBI connector.

[procyon82 · Accepted solution]

"If the number of sessions in the property over the given date range exceeds 500k sessions (25M for Premium)1, Analytics will employ a sampling algorithm which uses a sample set proportional to the distribution of sessions by day for the selected date range."

https://support.google.com/analytics/answer/2637192

I don't know if this means per query in Power BI (my guess would be yes).

This means you need to break down your queries and keep duplicating...not ideal, but it works.

http://community.powerbi.com/t5/Integrations-with-Files-and/Set-date-range-limit-on-google-analytics...