I’m trying to evaluate piwik for getting analytics on multiple kiosks.
I would like to track events and locations
my locations are in grouping, stat/city/site/location and events have about 4 parameters.
I am wondering how to effectively structure the data in piwik, so that I can get information across a single location, or an entire city. and get event counts grouped by type in various time increments such as yearly, monthly, weekly, daily, hourly, and weekly, daily, hourly averaged across a time-span
To achieve this, is it better to use the events as urls and location as custom variable, or vice verse? Are there any other options for this?
country
— An override value for the country. Should be set to the two letter country code of the visitor (lowercase), eg fr, de, us.
region
— An override value for the region. Should be set to the two letter region code as defined by MaxMind's GeoIP databases. See here for a list of them for every country (the region codes are located in the second column, to the left of the region name and to the right of the country code).
city
— An override value for the city. The name of the city the visitor is located in, eg, Tokyo.
lat
— An override value for the visitor's latitude, eg 22.456.
long
— An override value for the visitor's longitude, eg 22.456.
I need about 5 fields for location, and other than city, they all need to handle custom text.
Is there another set of values I can use to set up a tree structure for grouping the data? Maybe location information might work better as url and the events and related data as custom fields. I realize this is not the intended use of piwik, but I thought I might be able to use it in this case, rather than re-inventing the wheel for a marginally different use.
I think in simplest form, I want events to have three kinds of data
by date (hour,day,week,month,year)
by event data (event type,event sub-type, event argument, event id)
by location (country/region/city/site/terminal, where each item is string, instead of a lat/lon pair)
all three data items would act as cascading lists, allowing user to count, for example, how many times a category was looked up across all kiosks at a site week-by-week.