Yeah, I need to actually find the reference where I read the study (wasn't able to see it quickly so will probably look tonight) but IIRC there was a very clear distinction between "people using apps more and more" EXCEPT for news/reference in which case they were almost not using apps at all and only using mobile web (which certainly fits for my personal experience, I've found that every news or reference app I see ends up being more cumbersome and makes finding what I want slower then if I just used the browser and could move around more freely).

While it's certainly interesting (though not surprising) to see the level of entertainment activities I worry that attempting to "compete" time wise with twitter/facebook etc is not a very efficient use of our resources. It's better to make sure that when they DO want data they come to us (and that Twitter/Facebook etc are able to use our data well too :) preferably with a way to measure that but spreading that information is important).

James Alexander
Community Advocacy
Wikimedia Foundation
(415) 839-6885 x6716 @jamesofur

On Fri, Aug 28, 2015 at 12:53 PM, Tilman Bayer <tbayer@wikimedia.org> wrote:
Yes, good points, and of course those reports do not mean we have to
strive for the same ratios for our own traffic. One could imagine a
universe where the browser becomes "the app where I read Wikipedia
articles" for most people and we would still be happy with our
pageviews (if not with the state of the WWW in general). Still, the
two reports show that the environment has definitely been changing and
differs a lot between geographies - that's something to be aware of.

In any case, as we know, even US-based mobile Wikipedia readers still
overwhelmingly use the web version currently - actually, last month
the ratio of app pageviews among mobile pageviews was even a tiny bit
lower in the US than globally: 2.9% vs. 3.2%.

And conversely, the ratio of pageviews from the Global North is pretty
similar across all channels: 77% on desktop, 80% on apps, 78% on
mobile web.



Queries used for the above (all for July 2015):

SELECT SUM(IF (access_method = 'mobile app', view_count,
0))/SUM(view_count) FROM wmf.projectview_hourly WHERE year = 2015 AND
month = 7 AND agent_type='user' AND (access_method = 'mobile app' OR
access_method = 'mobile web');

SELECT SUM(IF (access_method = 'mobile app', view_count,
0))/SUM(view_count) FROM wmf.projectview_hourly WHERE year = 2015 AND
month = 7 AND agent_type='user' AND (access_method = 'mobile app' OR
access_method = 'mobile web') AND country_code='US';

SELECT access_method, SUM(IF (FIND_IN_SET(country_code,
'AD,AL,AT,AX,BA,BE,BG,CH,CY,CZ,DE,DK,EE,ES,FI,FO,FR,FX,GB,GG,GI,GL,GR,HR,HU,IE,IL,IM,IS,IT,JE,LI,LU,LV,MC,MD,ME,MK,MT,NL,NO,PL,PT,RO,RS,RU,SE,SI,SJ,SK,SM,TR,VA,AU,CA,HK,MO,NZ,JP,SG,KR,TW,US')
> 0, view_count, 0))/SUM(view_count) AS percent_GN FROM
wmf.projectview_hourly WHERE year = 2015 AND month = 7 AND
agent_type='user' GROUP BY access_method;



(FWIW, https://stats.wikimedia.org/wikimedia/squids/SquidReportCountryData.htm
has similar data, but is based on the old pageview defintion and
marked as unmaintained and unreliable there. It e.g. shows the iOS app
as having *only* users from the Global North ;)

On Thu, Aug 27, 2015 at 10:40 AM, Jon Robson <jdlrobson@gmail.com> wrote:
> It's worth noting that the 2nd graphic given on
> http://thenextweb.com/insider/2015/08/26/how-we-use-our-phones-is-changing-in-a-big-way-heres-the-data-that-proves-it/
> tells us that most app usage is on facebook, entertainment(games). I
> interpret this as the web is not as engaging as an app. Compare an
> addictive RPG mobile game with a blog post / wikipedia article - of
> course more time is going to be spent in the former. I don't think you
> can compare the two and a lot of these reports bother me since they
> seem to forget this.
>
> When I was commuting to school back in the day I invested my time on
> my Nokia 3210 on the game snake. When I got WAP I'd read BBC news sans
> images. When data was cheaper/faster - blog posts/news articles.  Now
> I have apps my options are endless.. Twitter.. Facebook... maps... ahh
>
> My interpretation is people are choosing to invest their internet time
> on things such as messaging and entertainment rather than web and that
> we need to find ways to engage people in the Wikipedia rabbit
> hole/experience, the platform not being relevant.
>
>
> On Thu, Aug 27, 2015 at 8:26 AM, Tilman Bayer <tbayer@wikimedia.org> wrote:
>> .. well, that would be a bit too simplified, but there's food for
>> thought in these two reports:
>>
>> http://qz.com/466089/the-fastest-growing-mobile-phone-markets-barely-use-apps/
>> ("In Asia and Africa, websites made up 90% and 96% of mobile
>> impressions, respectively, in the second quarter.
>> Their habits are a sharp contrast to the US, where apps accounted for
>> 91% of impressions. Globally, there’s a more even distribution, with
>> apps making up 56% of mobile impressions and websites comprising the
>> remainder.")
>>
>> http://thenextweb.com/insider/2015/08/26/how-we-use-our-phones-is-changing-in-a-big-way-heres-the-data-that-proves-it/
>> ("According to Yahoo’s data, the use of the browser on smartphones is
>> quickly declining [in the US]. From 2013 to 2015, the company saw the
>> usage time of the mobile browser drop by over 50 percent as they moved
>> to native apps instead.")
>>
>> --
>> Tilman Bayer
>> Senior Analyst
>> Wikimedia Foundation
>> IRC (Freenode): HaeB
>>
>> _______________________________________________
>> Mobile-l mailing list
>> Mobile-l@lists.wikimedia.org
>> https://lists.wikimedia.org/mailman/listinfo/mobile-l
>
>
>
> --
> Jon Robson
> * http://jonrobson.me.uk
> * https://www.facebook.com/jonrobson
> * @rakugojon



--
Tilman Bayer
Senior Analyst
Wikimedia Foundation
IRC (Freenode): HaeB

_______________________________________________
Mobile-l mailing list
Mobile-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mobile-l