I’m a Product Manager at Piwik PRO, now responsible for the maintenance and development of the Looker Studio Connector for Piwik PRO Analytics. In order to meet our client’s needs and expectations, we’ve just launched a project that aims to greatly improve the quality and usability of the Looker Studio data connector. To ensure access to information, transparency and to make our lives easier, I’d like to publicly share a list of known issues connected with Looker Studio, along with their statuses. I’ll be posting updates, adding estimated times of delivery when we set one. Thank you very much for all your patience and understanding!
Add the connector to the official Google Gallery and make updates smooth and easy (Q1 2024) (Change of authentication method required)
Known facts that are not bugs:
All metrics and dimensions appear as green fields in the Looker Studio connector for Piwik PRO - Google changed the way they treat data sources like Piwik PRO, and now lists both metrics and dimensions as “green” fields, which were previously reserved only for dimensions.
After dropping any metric listed as a green field into the “Metrics” definitions, it will change the color to blue and become a proper metric.
Hello Guys! I know there’s a lot of you waiting for an update here. We were working hard to meet the deadline and your expectations, and we’re finally ready to announce that a new version of the connector is ready. It will be published publicly on Monday, November 27th.
Here’s the list of changes that we managed to deliver:
Fixed Google Ads metrics/dimensions
Improved date/timestamp handling, added support for all date groups in reports like hour of day, day of week, week, month etc.
Fixed ‚Average page views in session’ metric
Fixed exclusion filters for is null/empty/equals to 0 conditions
Added support for exclusions with IN_LIST filters
Added handling complex conditions with IN_LIST filters (Example: Exclude Country in Denmark,Germany unless Device Type in Smartphone,Tablet)
Added support for all transformations for eCommerce metrics (Like Sum of product revenue (Orders), Average Revenue, Sum of product quantity, Median product quantity etc.)
Removed the product scope dimensions which are not available in Queries API and were breaking the connector (Like Product revenue (Array))
Fixed bounce rate showing wrong numbers.
We’re not stopping here - the next versions will now be released regularly. Stay tuned!
Hello! A new version of the Looker Studio connector is already here
The final changelog:
Fixed Google Ads metrics&dimensions
Improved date/timestamp handling, added support for all date groups in reports like hour of day, day of week, week, month etc.
Fixed ‚Average page views in session’ metric
Added filters for ‘is null’/empty/‘equals to 0’ conditions (for both metrics and dimensions)
Added support for exclusions with IN_LIST filters
Added handling complex conditions with IN_LIST filters (Example: Exclude Country in Denmark,Germany unless Device Type in Smartphone,Tablet)
Added support for all transformations for eCommerce metrics (Like ‘Sum of product revenue (Orders)’, ‘Average Revenue’, ‘Sum of product quantity’, ‘Median product quantity’ etc.)
Removed the product scope dimensions which are not available in Queries API and were breaking the connector (Like ‘Product revenue (Array)’)
Hello! I’d like to drop a little holiday gift and share a new Looker Studio connector version 1.2.0 with you. Here’s a list of changes we managed to introduce:
Added support for filters with ‚Equal to (=)’ and ‚Contains’ operators for enum type fields (Disclaimer - Location dimensions (Country, City) will work like ‚Starts with’ filter when the ‚Contains’ operator is chosen in Looker Studio. This limitation will be resolved in Q2 2024)
Added ‚Goal ID’ and ‚Goal UUID’ dimensions (Disclaimer - Those fields support only the ‚equal to’ operator due to API limitations)
Fixed pie/donut charts showing wrong numbers
Fixed calculations after changing the number of pie/donut slices in charts
Fixed using dimensions like ‚Country’, ‚Browser name’, ‚Device type’ with pie/donut charts
Fixed Breakdown dimensions breaking the connector
Fixed filtering by Website name (and other enum type fields) in Meta sites
Added support for more than 1000 websites/meta sites in the Site selector
Added limit of 100 results for enum type fields.
Additional important info for the above change log (which will also be included in the official connector manual):
The RegExp and ‚Starts with’ (which in fact is also a RegExp) filter types are still not supported. Our Analytics team is running research on the possibility of implementing it in our API, which would open the way to add it in Looker Studio.
Please, bear in mind that filters in Looker Studio are case sensitive (so to reflect some operators marked as ‚case insensitive’ in Piwik PRO UI, you’ll have to add additional conditions in Looker Studio).
When using filters on enum type fields, we have introduced a limit of 100 results. This performance enhancement helps to provide faster load times and a more efficient browsing experience. If you encounter a message indicating that too many values matched the filter, please refine your search criteria to continue.
And last but not least, this is another batch of fixes but we still have some points in our backlog that we’re going to take care of in the upcoming releases. If you find something that is still not working properly and it hasn’t been mentioned in this post, please report it here in the community forum or to our support.
We would highly appreciate also positive comments, so if you have any thoughts on the new connector, don’t hesitate to share.
Added API token caching to reduce the number of token requests
Added support for specific goal metrics (instead of filtering out a specific goal while looking for example at the 'Goal conversion rate’ metric, you should choose a metric 'XYZ goal conversion rate’)
Added support for ‘Unique count of’, ‘Min’, ‘Max’, ‘Average’, ‘Sum of’ and ‘Median’ transformations for all eligible columns
Improved calculations of some aggregations for scorecards (they’re now less prone to user errors)
Added support for RegExp Contains (RegExp partial match) and RegExp Match (RegExp exact match) filters (Note - Those filters are not supported for enum type columns (like Country, City etc.))
Fixed filtering for transformed fields (e.g. Page Path).
You should also be informed that we are actively working on bringing the connector to a state where all Google’s requirements would be met and we could publish it as an official connector in the Looker Studio Gallery. This would allow for more smooth updates. Now, you should still remember to follow this instruction for each of your dashboards if you want to start using the new version of our connector.
I would also like to emphasize that Looker Studio users should be aware that in complex dashboards there might occur errors caused by exceeding the limitation put on Looker Studio by Google. The limitation is for simultaneous request executions per user and we have no influence on that limit. As a solution, Google provided the following recommendations:
I hope you’re going to enjoy the new features. We would appreciate your feedback and suggestions for further improvements
Does this mean that I have to redo all of data connections in Looker? Most of my Piwik Pro data sources have custom metrics or dimensions. Will I have to redo those?
There is still a massiv error using Looker with the Piwik connector. Still the revenue differs from Piwik and Looker. On a yearly scale I am missing 4 million € that are not shown. I just updated the connector 10 min ago. Best part … the bar charts show higher revenue than the overview boxes … still not the correct value.
New version 1.3.1, and still visitors total in Looker and Piwik is different:
same range, no filter, looker shows 80K visitors while in Piwik only 65K.
Version 1.2 at least showed the same value as Piwik interface.
Hi @muttley90, you are correct - we have not yet updated this behaviour as it will most likely require some changes to the core product and its API. Sorry for the inconvenience. I will update this thread when the fix is released.
In the meantime, you can use the “count distinct” aggregation on visitor ID:
Is it only me or is there a problem regarding goal conversion rates? I have an account with two goals. Looking into conversion rate in piwik it’s for 3 days of testing 0.06% and when looking in LS it’s 0.16% (same amount of sessions and conversions both places when I look at those numbers). Any suggestions to why?
The 1.3.2. and 1.3.3 versions introduced changes to the querying mechanism for some of the metrics/dimensions available in the connector. This should resolve issues with unnecessary aggregations (e.g. Visitors, Goal conversion rate, Bounce rate) and therefore discrepancies between LS reports and Piwik PRO Analytics UI for these cases.
I’m running into issues with the dimensions “Goals converted in session” and “Goals converted in session (uuid)”. The “Goals converted in session” dimension always seem to give an error (ID e407b1cd) and the “Goals converted in session (uuid)” dimension massively underreports the amount of sessions compared to the goal report in Piwik.
Would you mind sharing a screenshot of the Piwik PRO report that you are trying to replicate/comparing Looker Studio to? Either here or in a private message.
and share a new