[OSM-talk-be] some osm.be statistics

Mon Nov 2 19:47:04 UTC 2015

Hi all,

I promised the Mapper of the Month team I'd have a look at some statistics
way too long ago. It being a kind of birthday now, here they are. Data is
collected by Google Analytics.

*Basics*

There are 74.000 pageloads in total, but only 32.000 of those are from
something else than the front page. There are about 30.000 unique visitors
so far. About 8000 visitors were seen more than once. (of course, Google
doesn't know -everything- we do, so some people are likely counted several
times)

On a monthly basis, it looks like this:

http://imgur.com/cv3W6Iv

So clear growth, however september and october aren't very impressive.

*Language*

21% of language-identifiable pageloads are on French language pages, 79% is
Dutch. To reflect Belgium, that should really be 40/60. It doesn't seem to
have to do with content; the statistics stay the same if you just look at
the main pages.
If you take the browser language, there's still 80% Dutch speakers.
However, there is only 7% French speakers, with another 12% English and 3%
various other languages. I tend to assume that relatively more Dutch
speakers would have English language browsers than French speakers. But
maybe that's no true for people interested in OSM?
Anyway you look at it, the website is clearly underrepresented in the
French speaking area.

*Looking at what people look at.*

Something I absolutely don't understand is that among the French speakers,
the most popular page is the French index. This has almost as much visitors
as the Dutch index. For other pages, it's the more usual 80/20 devision.
Maybe some Dutch speakers got to the French index by mistake, then change
language?

Another surprise for me was that by far the most popular of the main pages
was the usage policy (red in the graph below). So that would be a good
place to put more effort, I guess. Maybe have a comment section there, to
see if they are missing anything?

It's not as easy as I thought identifying the Mapper of the Month series.
Some visits are identified by Analytics, others probably saw it through the
blog section and are not seperatable. Maybe the images have seperate load
statistics to have a better measure?
Anyway, you can group the "more dynamic content" as the top most three
categories on the graph below.
You can see that among the French speakers, that makes up 30% of all
pageloads, while for the Dutch speakers, it's more like 15%.
I don't know what kind of conclusions that could entail.

http://imgur.com/egiu5Lr

*Where did they come from*

Most visits come from Google, 66% of the total. 27% typed a url or clicked
a favorite. Less then 5% clicked a link, and only 1% came from the social
media.

Biggest amongst the links are the ones from escada's diary entries for the
Mapper of the Month.
OKFN and wiki.osm are other obvious links. Less obvious are the 100 visits
coming from a post by AS adventure,
http://www.asadventure.com/benl/content/nl/alles-op-een-kaart-met-je-gps .
Which IMHO goes to show that we should try to get more niche users like
that link to our page.

*On Mapper of the Month*

Considering that the English language posts on Escada's diary get an
exposure of about 3000 views each (guestimate, based on the loads of the
imgur images I unclude in my posts), the Dutch and French translations on
the website get quite a low number of views.

In the best months, the project gets about 120 views, and some months it's
really very low.

http://imgur.com/oOzNvGO

However, I don't think we can trust these statistics. I used the page title
here, and I have the impression that only gives the good title when you are
following a direct link.The same goes for using the page URL. Yes, yes,
Google Analytics, lets you do all sorts of flashy stuff. But basic
statistics, forget it.

Here I took all the Blog traffic together. You can clearly see a jump when
the project was started, but still not that impressive numbers of views.
The three high months are because of the articles on the Missing Maps and
Mapping Public transport.

http://imgur.com/py054Ze

So unfortunately I don't think we can really quantify the Mapper of the
Month. Here's another simple approach I tried to measure "community
engagement" with the articles. There was some discussion whether the
articles should be shorter or that long is good. So I counted the words of
all the mapper interviews, and the number of responses to the Diary posts.
Just a manual review, nothing fancy. I know, that's kind of silly, but it's
the best I could think of.

In the graph I plotted the number of words horizontally, the number of
comments vertically. The blue dot represents one mapper of the month
interview. There are to little interviews to really speak of a trend, but
you definitely would not read conclude here that shorter articles get more
feedback.

http://imgur.com/au1cjP7

Which makes sense to me. If you're not that interested in OSM or the
interviewee, you're not going to read one of these articles. But if you do
find things like this interesting, chances are you just can't get enough.

If anyone is interested, I can send you the messy excels I used. Analytics
allows for a public dashboard, so if you want to keep up to data, just ask
me what you would like to see constantly up to date. Or if you want access
to the Analytics website, get in touch with Ben Abelshausen.

-- 
Joost @
Openstreetmap <http://www.openstreetmap.org/user/joost%20schouppe/> |
Twitter <https://twitter.com/joostjakob> | LinkedIn
<https://www.linkedin.com/pub/joost-schouppe/48/939/603> | Meetup
<http://www.meetup.com/OpenStreetMap-Belgium/members/97979802/> | Reddit
<https://www.reddit.com/u/joostjakob> | Wordpress
<https://joostschouppe.wordpress.com/>

-- 
Joost @
Openstreetmap <http://www.openstreetmap.org/user/joost%20schouppe/> |
Twitter <https://twitter.com/joostjakob> | LinkedIn
<https://www.linkedin.com/pub/joost-schouppe/48/939/603> | Meetup
<http://www.meetup.com/OpenStreetMap-Belgium/members/97979802/> | Reddit
<https://www.reddit.com/u/joostjakob> | Wordpress
<https://joostschouppe.wordpress.com/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-be/attachments/20151102/e927aa6f/attachment.htm>