Two years ago we built and released a Google Analytics feature for Moodle which offered some robust reporting for Moodle – beyond the standard url logging it logged the breadcrumb from Moodle – see here for details. This was released as a plugin on the Moodle plugin directory.
David Bezemer of UP Learning took this concept/plugin and then expanded it to included Piwik an open source analytics system. The plugin is also available in the plugin directory.
Using this plugin requires you to install your own copy of Piwik separately, which you can use for all your websites. So I used it for the Moodlemoot Edinburgh this year, and below include some of the reports that were possible to generate from it.
Installing the plugin
The installation was very simple. I went to the Moodle plugin directory and was able to download the correct version for the Moodlemoot Moodle site.
After the install, there was a selection of new settings required which come from the Piwik install of the analytics site itself.
local analytics settings
So I configured it for the Piwik install I had access to, and that was that.
Using the plugin
The plugin itself didn’t have any features inside of it. It just worked adding tracking into the page so that the Piwik platform held the information. Unlike Google Analytics, this is then totally your own on your own servers rather than using a 3rd party system. So it really comes down to the features and what it can track/report on which I deal with now.
In many ways, Piwik is a lot like Google Analytics in features with a good dashboard and dril down information available. Below are some of the reports but I will start with the first one which uses the page title as the dril down like we did on the G:A plugin, so that you can see categories of courses and course data collated together. All these reports are from different time-spans during the Moot.
Page Area Expanding
Each image takes a progressive expansion.
Course List in a Category
Activity type in a course
Specific Activities of a type in a course
So you can see that it is possible to view the data for a Moodle Course Category, then the course, then types of activities and then specific activities.
You can also see how the usage changed over time for any row – which it calls evolution and compare them with another section too. So two courses as in the below example\
Two courses evolution compared
This can be useful for so many ways.
- Comparing courses over time in a category
- Comparing categories over time against each other
- comparing different resources in a course to see how their usage changes over time and so on.
I cant list all the valuable information this can provide for an organisation except, it is mind boggling.
So what about other reports?
Here is a collection of some of the other reports all equally useful for different reasons:
Visits over time is useful to see when the site is quietest for maintenance windows.
visits- per local time
Visit duration is nice to see how long people stay on the site, course etc.
Of course not all data is just graphs, you can get tables of data in a list, graphs or word clouds. Below is the same data in Piechart and cloud.
pages per visit
pages per visit cloud
You can get breakdown for countries and continents:
Browser and Device information is super interesting too and helps understand your users. Such as which browsers you need to focus testing on, which devices – how many are using on mobile and so on.
browsers full data
mobile vs desktop
Another interesting aspect, is the visitor log itself. You can view the log for visitors and also for repeat users (you dont get a name / userid – just their IP and a uniquely assigned Piwik id). I have removed the IP and some IDs from the images for privacy purposes.
In the full profile, you can see the user had 5 visits totalling 3 hours 41 minutes on the site.
Example Visitor Log Entry
Example Visitor Profile
There are more reports, but just wanted to include these to give you a flavour of what is possible. I use Google Analytics on pretty much every site I run and advise others to do so as well because of the value of the data.
However, for most of those Piwik is probably more than sufficient to use and preferable to quite a lot who tend to balk at using cloud solutions like Google, Microsoft etc.
So I see this being adopted in this summer by many of the organisations I have talked to who want to keep this data inhouse and still get quality reporting.
Great job Piwik – and thanks David for taking the work Bas Brands did and adopting to Piwik.