In order to better understand geohack usage patterns, instrument GeoHack ([[ https://bitbucket.org/magnusmanske/geohack/src/master/public_html/?at=master | code ]]) to collect the following data:
* visiting geohack page (count per wiki, per monitor height - in pixels?)
* interacting with the embedded map, e.g. pan/zoom/click. (count per wiki, once for each type of action)
* clicking on a link that contains {londegdec} or {latdegdec} or other similar geohack template parameters. Ignore {pagename} and similar as that could be a link back to wikipedia (count per wiki, per target domain of the link)
* clicking on a link that does not contain any of the above {...} values, e.g. edit link, explanation links, etc. (count per wiki)
* clicking on a language switch (count per wiki)
* count whenever user simply goes back or closes the page, without making any actions.
Our KPI questions are:
* How many people visit geohack page and are happy with what they see (do nothing)
* Out of those people, how many people interact with the map (if available), and do nothing afterwards, vs how many people simply click on an external map link, vs do both