A new site traffic reporting service for OpenEdition communities
A new traffic reporting service has been rolled out for all sites across OpenEdition’s four platforms. The data reports are publicly available and can be consulted at https://analytics.openedition.org/. This service enables OpenEdition communities to assess the traffic to a site over time, as well as analyse visitor behaviour and reading habits in depth. The data is collected using Matomo, an open source tool for analysing visitor traffic. Jean-François Rivière, head of the data team at OpenEdition, and Emilie Cornillaux, document information systems administrator for the data team, tell us more about this new service for OpenEdition users.
- OpenEdition had been using the AWStats analytics tool to measure traffic since 2005. Why switch to Matomo?
The open source AWStats software we had been using is a straightforward and efficient tool offering a wide range of statistical reports. Like AWStats, Matomo is also an open source tool for collecting a set of data which can then be used to analyse website traffic. However, we find it much more user-friendly than AWStats. Its interface is more modern and it offers a large number of widgets and extensions. You can tailor it more to your needs and it makes analysing the various reports generated easier, especially as it offers a variety of display modes and export formats. On top of that, the CNIL (French data authority) recommends it and it complies with the General Data Protection Regulation (GDPR). That’s why we chose this tool.
- What were the different stages involved in rolling out the tool?
We wanted to be able to gather data from all OpenEdition websites, which meant creating 8,000 websites in Matomo. The first major task was to understand how the Matomo API works. We needed to ensure that existing sites, as well as any new sites published in Lodel and WordPress, would be automatically created. The second important step was to configure the Matomo instance to optimize its use and reduce the waiting time, comply with the GDPR, and take all possible scenarios into account (keeping count of proxied URLs, excluding URL query parameters that are not relevant to data analysis, limiting the type of files considered during downloads, etc.). The final step involved adjusting the interface settings to determine which reports users could consult by default using the widgets offered by the dashboard, and identifying the extensions to be installed and those to be deactivated according to the agreed requirements.
- What challenges did you face? Did each platform have its own specific needs?
The main technical challenge we faced at OpenEdition was the sheer number of sites, as always. You cannot handle data for one site in the same way as for 8,000 sites. You need to take data size into account, automate processing as much as possible and anticipate performance issues.
As for the specific needs of platforms, the interface is the same for all users in Matomo, with a dashboard for each site. Calenda is a bit different as the platform consists of a single website. The statistics are therefore available for the platform as a whole and not for each individual event.
- The statistics can also be viewed directly from the site administration interface. Do the site administrators have access to specific features?
The links provided in the administration interfaces of journal and blog sites provide direct access to the Matomo dashboard for the site concerned, along with the corresponding statistical reports. Given that the statistics are publicly available, site administrators have access to exactly the same features as other users. The administration interface simply avoids the extra step of searching for the site on OpenEdition.
- What about the other data reporting services offered by OpenEdition?
Rolling out Matomo is part of a more global strategy geared to developing data reporting services.
In January 2021, we set up a Counter 5 usage reporting service aimed at institutions that have signed up to the OpenEdition Freemium program. It enables them to analyse traffic to the OpenEdition Journals and Books platforms from their campus or via online tools provided by their library.
This service is based on log analysis using the ezPAARSE software tool. It has replaced our preceding Counter 4 service, which was based on AWStats.
Along the same lines, in the near future (2023), we are going to replace the current usage measurement service based on AWStats with a new service, also based on ezPAARSE. This service is available to publishers on the OpenEdition Journals and Books platforms. It enables them to obtain traffic reports for individual documents (e.g. the number of views over a given period for each article, book or journal issue number, and/or format). The new version of this service will also apply the Counter 5 standard, meaning that the statistics available to libraries can be correlated with those available to publishers.
Once we have rolled everything out, our offer will include a Matomo reporting service for OpenEdition’s four platforms and two services based on ezPAARSE and Counter 5 for the OpenEdition Journals and OpenEdition Books platforms, designed for both publishers and subscribing institutions.
Resources and documents for our users (in French)
Find out more about Matomo and web traffic analytics: