Piano Analytics and Google Analytics (or Universal Analytics) are tools with completely different calculation logic, which makes it difficult to compare the collected data. It is therefore not possible to have strictly the same data on these two tools, although it is possible to explain some differences between Piano and Google.
Fundamental discrepancies
First of all, it is essential to bear in mind that Google is sampling your data, while Piano returns all the collected data (by thinking to deactivate the eco mode). Also, the metrics are not necessarily calculated the same way.
Note that GA4 estimates the number of session IDs to calculate the total number of sessions, as explained in this Google article.
On Piano Analytics as on Google Analytics, the visitor's identifier is stored for 13 months, by default. The first event generated by a visitor will result in the creation of a visit, which may be over:
- after 30 minutes of inactivity
- at midnight
- if the visitor identification cookie is deleted
However, with GA3 (Universal Analytics), a visit also ends when new campaign parameters are found, resulting in the creation of a new visit, which is not the case with our tools. With GA4, the source of a session depends on the scope of the dimensions analysed, as explained in their online documentation.
Sources discrepancies
Although, fundamentally, the main Analytics metrics are almost calculated the same way, it is common to find differences in volumes between the sources.
As explained, on GA3, when new campaign parameters are found, a new session is generated, thus valuing the marketing sources. Moreover, Google's attribution model (last touch non-direct) is not the same as Piano's. For example, a session starting with an organic source may be assigned to a marketing source instead, if such a source has been encountered in a previous period, as explained in their Flow Chart.
Regarding GA4, the result will depend on the scope of the analysed dimensions.
On Piano, the source of the visit is unique and defined at the beginning of the visit. If a visitor arrives from a marketing source during an already existing visit, it will generate a click.campaign.offsite event (see this documentation), but will not create a new visit and therefore will not value the marketing source in question.
Tagging
You also need to ensure that the tracking on the two platforms is the same: Are the events triggered at the same time? Are they subject to the Privacy in the same way?
Piano Analytics allows you to benefit from the ePrivacy Exemption, which is not the case with Google Analytics. In the case of a GDPR compliant tagging, no Google event is generated until the visitor has accepted the cookie banner, whereas Piano events can be generated, even if the CMP is denied, although they are not supposed to collect any marketing source in exempt mode. If you benefit from the exemption tagging, then you will need to check whether you have exempted marketing sources or not, in which case this could explain a higher number of marketing visits to Piano.
It is also important to ensure that, in general, the site is tagged the same way. If certain pages are tagged using only one of the two tools, this could make comparisons more difficult.
UTM
You also need to verify that the campaign tagging corresponds to the expectations of each tool. Our new SDKs are fully capable of interpreting Google's UTMs (see this documentation), but our processing chain requires, as a minimum, that the src_medium and src_campaign properties are set to take account of the marketing source of a visit. So, if you use UTM tracking, but your links do not contain the utm_medium and utm_campaign parameters, the marketing sources of your visits will not be taken into account.
As a solution, you can create processing rules that associate the UTM parameters of your choice with the native sources properties of your choice, allowing you to customise the assignments, as suggested in this documentation.
You should also bear in mind that the Piano properties dedicated to collecting UTMs are event-scope properties, which can collect several values per visit. It will therefore be necessary to analyse our native sources properties (Source, Campaign - Name, Campaign - Type, etc.) if you make comparisons with Google's tools.
Other parameters
However, Google campaigns based on URL parameters such as the gclid cannot be understood by our tools, as it is a chain generated on their own.
If you are in a position to know which gclid should be associated with which information, you can create matching tables to be integrated into the processing rules of Piano's native source properties.
To resume
| Unique source | Complete data | Exempted visits | at_ | UTM | gclid | |
| Piano | ||||||