I want to thank everyone for this robust discussion! We appreciate the
feedback very much. I want to recognize the role that Joseph Seddon has
played in encouraging us to share more results with this email list; he's
the #1 reason that I started this thread.

I also want to thank Peter Coombe, my teammate on the banner crew, for
suggesting so many of the notes that I am about to provide. His insight
into mediawiki and the more mysterious parts of centralnotice are a
constant value.

Lots of responses:

   - I received a note off-thread re: bad wrapping of the Read/edit/more
   links between 920 - 1060px. We'll take another pass at that.


   - Agreed about the Close X. We'll make that clearer and easier to click.


   - This is a good opportunity to reiterate that the norm for us is to
   only show a single 'large' banner, per device, in a given fundraising
   campaign. It usually serves as a reader's first banner impression, converts
   extremely well, and then we move to 'small' banners. Sometimes we test
   alternate sequences, but this is normal for the vast majority of readers.


   - This banner is a desktop banner; it will not be shown to anyone on the
   mobile site. However, it is possible for you to follow the preview link I
   provided on any device. If you were viewing on a large phone or tablet, it
   might have seemed particularly strange.


   - Gergo Tisza, issues with the Timeless skin will not be a problem,
   because this banner will only be targeted to logged out users who can not
   pick a custom skin.


   - Petr Kadlec, this work-in-progress banner is using some copy written
   in English, as well as a couple translation templates; the final version
   will all be hardcoded English. However, your bug sounds strange - could you
   try opening the link in an incognito browser and, if you're still getting a
   problem, send me a screenshot?


   - Samuel Klein, that red underline is a pretty consistent element in our
   designs. For what it's worth, we actually tried a version where the
   underline did function as a hyperlink, but it had no effect on giving rate.
   Thanks for your other testing notes!


   - Additionally: we run surveys to try to understand reader and donor
   views on how obtrusive and frequent our on-Wikipedia fundraising campaigns
   are perceived to be. We consistently see feedback that the quantity of
   banners is "about right," and that the banners are, to most people, not
   overly aggressive.

We'll all be monitoring this thread; thank you, again, to everyone who took
the time to respond. In all honesty, I am mostly un-circumspect. I believe
that we have to take risks and try many new ideas; wins, and losses, in our
work are both frequent and often surprising. I'd point to a blog post we
published recently
<https://blog.wikimedia.org/2017/10/03/fundraising-banner-limit/> that
shows the diminishing returns of banner impressions over the course of one
of our campaigns. It is not true that we could simply run more banners, for
longer periods of time, and expect the same returns.

It's a constant process of experimentation and I appreciate having such an
engaged community to review, improve, and motivate our work.

regards,
sam

On Wed, Nov 15, 2017 at 8:23 AM, John Erling Blad <jeb...@gmail.com> wrote:

> What kind of reasonably new device has that kind of resolution?
>
> On Wed, Nov 15, 2017 at 12:41 PM, <jay...@gmail.com> wrote:
>
> > Patrik, try it on a 640x480 screen. :P
> >
> > On Wed, 15 Nov 2017 16:06 pajz <pajzm...@gmail.com> wrote:
> >
> > > >
> > > > This test would run for 1 to 2 hours, and then we'd evaluate results
> to
> > > see
> > > > if it's worth spending any more time on the concept. For now, we're
> > > simply
> > > > hiding the banner all together below 920px, as at smaller viewports
> it
> > > > begins to interfere with site navigation elements.
> > >
> > >
> > > It is truly an amazing banner, as it might be the first donation banner
> > in
> > > the history of donation banners that actually improve the readability
> of
> > > our content. Indeed, we have never been closer to line length
> > > recommendations for optimal readability than with this banner.
> > >
> > > Patrik
> > >
> > >
> > > On 15 November 2017 at 01:25, Liam Wyatt <liamwy...@gmail.com> wrote:
> > >
> > > > >
> > > > > TL;DR: A short test of a new banner concept will help us decide if
> > it's
> > > > > worth iteration and improvement.
> > > > >
> > > > > Here's a link to the banner:
> > > > > https://en.wikipedia.org/wiki/Apple?banner=dsk_p1_lg_
> > > > > right10&country=US&force=1
> > > >
> > > >
> > > > I appreciate that you're trying to innovate, and this requires
> pushing
> > > > boundaries. I also appreciate that the WMF fundraising principle
> > "minimal
> > > > disruption"[1] is subjective thing - and that the WMF has
> traditionally
> > > > taken this to mean "be disruptive for the shortest period" rather
> than
> > > "be
> > > > less disruptive overall, even if it takes longer".
> > > > However, I do feel that *resizing all the article content to the
> left,
> > > and
> > > > occupying the entire right side third of the screen with a donation
> > > request
> > > > that follows me as I scroll *is significantly more than minimally
> > > > disruptive. Personally, I would request that this design approach not
> > be
> > > > pursued on that basis.
> > > >
> > > > -Liam / Wittylama
> > > >
> > > > p.s. Also, on my screen at least, the "x" dismiss button in the
> > top-right
> > > > is so small that it hides under Chrome's translucent scrollbar
> > > >
> > > > [1] that https://meta.wikimedia.org/wiki/Fundraising_principles#
> > > > Fundraising
> > > > _______________________________________________
> > > > Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/
> > > > wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/
> > > > wiki/Wikimedia-l
> > > > New messages to: Wikimedia-l@lists.wikimedia.org
> > > > Unsubscribe: https://lists.wikimedia.org/
> mailman/listinfo/wikimedia-l,
> > > > <mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe>
> > > >
> > > _______________________________________________
> > > Wikimedia-l mailing list, guidelines at:
> > > https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and
> > > https://meta.wikimedia.org/wiki/Wikimedia-l
> > > New messages to: Wikimedia-l@lists.wikimedia.org
> > > Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,
> > > <mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe>
> > _______________________________________________
> > Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/
> > wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/
> > wiki/Wikimedia-l
> > New messages to: Wikimedia-l@lists.wikimedia.org
> > Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,
> > <mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe>
> >
> _______________________________________________
> Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/
> wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/
> wiki/Wikimedia-l
> New messages to: Wikimedia-l@lists.wikimedia.org
> Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,
> <mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe>
>
_______________________________________________
Wikimedia-l mailing list, guidelines at: 
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and 
https://meta.wikimedia.org/wiki/Wikimedia-l
New messages to: Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, 
<mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe>

Reply via email to