Looker Studio connector: specifics and new features – Piano Support

This article details several of the connector's specific behaviors, and highlights some of the new features. You'll also find explanations of the data discrepancies between Looker Studio and Piano Analytics.

There are different reasons for data discrepancies between Piano Analytics and Looker Studio, the connector being limited by the tool.

API behavior

Metrics such as "Visits" or "Visitors" are not summable metrics. The "TOTAL" result line displayed by our tools shows the deduplicated result, so as not to count the same visit or visitor several times on a given time period, as detailed in this article.

However, Looker Studio is not able to deduplicate this information, since the tool doesn't know the calculation we perform to proceed. So it simply sums up the lines. It is therefore likely that the tool will display larger values than our tools.

API limits

Looker Studio does not allow our connector to freely query the data. Thus, a query created on Looker Studio cannot exceed 10,000 rows of results. You can check the number of rows returned by a query by clicking on the appropriate button in the bottom right-hand corner from the Data Query interface:

If your query exceeds this limit, you will need to modify it or export the data from our tools and import it into Looker Studio.

The connector is also subject to our API v3 limitations, detailed here.

A looker board user can make no more than 5 concurrent calls to our API (1box=1call), and this limit rises to 20 concurrents calls across the customer's organization. However, the connector has a refresh system to ensure that your data is retrieved quickly.

It is possible that you may get errors of this type:

Please note that these errors are specific to Google, and not to our API. We recommend that you contact Google support, or possibly reconnect your data source.

Filters (New feature)

Although the 10,000-row limit may be restrictive in certain cases, there is a more frequent need to analyze query results with one or more filters applied.

With the new version of the connector (data source "Piano Analytics"), a significant improvement has been made to filter management. Previously, filters applied in Looker Studio were only considered on the first 10,000 rows of results returned by the Piano Analytics data query, which could lead to discrepancies between the results displayed in Looker Studio and those seen directly in Piano Analytics.

Now, with this update, filters are applied directly at the API call level, ensuring that the entire dataset is properly filtered during extraction. This allows for consistent results across both platforms and improves the accuracy of reports created in Looker Studio.

All of your queries with more than 10,000 rows of results will now be properly filtered.

Segments (New feature)

No need to create custom metrics to segment your analyses anymore! With the new version of the Looker Studio connector, you can now import your Piano Analytics segments directly into your report widgets. The new version features a global segment selector, available during the setup of the "Piano Analytics" data source, as well as a custom segment selector accessible from the widget configuration menu.

Global segment: Applies the segment to your entire dashboard.
Custom segment: Applies the segment only to the selected box.

Sort

The sorting applied by Looker Studio may differ according to your request. It is not possible to choose the sort to be applied at the time of the API call, as is the case with our tools.

If you load a query that exceeds 10,000 rows of results, the first 10,000 rows returned may be different from those displayed by our tools.

Eco-mode

As with data comparisons between our tools, or our API and our tools, you'll need to ensure that the eco-mode is deactivated. This mode can lead to a maximum 2% difference in your results, in order to reduce the environmental impact of your queries.

By default, it is not active in an API call, and therefore has no impact on the results returned by Looker Studio.

Boolean and Date typed properties

Boolean properties are not compatible with Looker Studio's AT Internet connector.

If you wish to upload this information to Looker Studio, please contact an administrator to create a custom property as shown below, converting your search_same_day property into a string.

However, these modifications are not retroactive.

The date and time properties from Piano Analytics are also reprocessed using Looker Studio's own function, sometimes in a different format.

For this reason, we've provided an example of how to make the 'Date' property work. The steps are reproducible for any time-stamped information present in our events, you'll just need to modify the rule to take the right part of the information depending on what you want to reproduce.

The idea is to create a new custom property that collects the date parameter from hits and retrieves only the month in string format. This way, the property won't be reprocessed once in Looker.

First, create your custom string property (let's take Looker_month as an example) and validate your draft:

Go to the configuration menu for this property, then Processing and finally Snippets, and create this snippet:

Then create the rule:

Before step 3, the property contains the month numbers (01, 02...) and you can add a mapping table to get the real names of the months like this :

You can then simply use this property in Looker Studio to obtain the correct month information.

Please note, however, that these string-format properties cannot be used in Looker's time charts, as the latter necessarily use a date format. However, our native date format properties may work for this type of chart.