As a Product Manager, I'd like to know how much traffic, on average, goes to the wikimedia.org portal page, so that I can decide how to prioritise improvements to that portal.
Description
Related Objects
Event Timeline
We have the raw numbers (http://discovery.wmflabs.org/portal/#pageview_tab) Do you want portal traffic as % of overall traffic to all WMF sites?
@mpopov Well, this task seems to be about wikimedia.org rather than wikipedia.org.
@debt Was the above intentional? If so, we should consider whether it's worthwhile actually doing this analysis. The wikipedia.org portal is all that's really in scope at the minute for Discovery; see T110070#1653320 for background on that.
Yup, it is intentional. One recent commentator to our new wikipedia portal update was to ask if we would do something similar to wikimedia.org. :)
I kind of just wanted a ballpark figure...nothing special.
I think producing a quick, one-off report would be fine. We can use this task for that; I'll adjust it accordingly.
That said, although I am open to the data changing our minds, I expect we will maintain our previous commitment to only focus on wikipedia.org. This makes this task fairly low priority.
Let's add in another site, just in case: www.wikipedia.nl, as referenced by this: https://office.wikimedia.org/wiki/ToDo/Chapters_coordinator/Reports/Week_2007-44
This is becoming a bit of a moving target. I'm not comfortable with that. Given our restrictions in resourcing both in the engineering and analysis front, the analysis should be restricted to wikimedia.org and wikipedia.org.
Running a script that's counting pageviews from the past 28 days. Might take a while.
@mpopov will check whether the script is done, and move this task to the correct column accordingly.
@debt: Do you want a report-report? I don't know what to say that isn't already clear from looking at the charts.
Hi @mpopov thanks for the work there.
It's a bit weird, because I didn't expect that many people to be hitting the page at wikimedia.org. But, maybe a lot of them could be because of typo's ("m" instead of "p") or simple curiosity of the page.
I'm good with this being done-done!