Wikipedia’s annual fundraiser is in progress. If you haven’t noticed already, when you go to the site, there’s a banner on the top that asks for donations. A few weeks ago, Wikipedia tested four different banners (below) to see which one resulted in the most donations, and they just posted the data for the test (along with some others). Can you visualize this?
The Jimmy appeal banner was the big performer by far, earning more donations as wells a much higher click through rate. The results page shows about twice as much donations for the Jimmy banner, but via the spreadsheet with the more specific breakdowns, that’s actually because the Jimmy ad ran twice as much as the other three.
What’s your take on the data? Get the spreadsheet here, and post links to your graphics in the comments.
Deadline: Friday, November 19.
So? They made a color banner with an important sounding message. That will obviously get a higher clickthrough rate than a silly quote about a brain massage. Someone out there finds that surprising?
Does the data have to be surprising to be a useful exercise?
Surprising? no. But it’s certainly not useful.
I disagree. Sometimes text ads perform better than display. Sometimes smaller ads perform better than big ones. With these numbers, they can be sure, where even small percentages can make a big difference at Wikipedia scale.
Additionally, this is one study out of several. And the numbers above are only a few from more comprehensive metrics.
And ultimately, yes, still a useful exercise for FD readers to get some practice and to learn from others.
Not at all, I agree. Only a fool would rely on assumptions when there is this kind of money at stake.Your assumptions may be correct, but data like this is the only way to be sure.
Three of them are clever, one is authentic. The clever ads are obvious attempts at marketing, the authentic ad is a message from one real person to another real person. The results are clear about which approach works better for this audience.
Right! I think the words “personal appeal” did the magic.
I think it was the sexy picture of Jimmy.
Information Is Beautiful already got there:
This is a great post about the problems with David McC’s viz. What I think he did (and me, too) is take aggregated data from the Wikimedia page, not from the source spreadsheet. The author of this post has done a great job on really looking into the data.
(In my defence, my viz was really only designed to show how a linear representation of the data is better than a box)
worth looking at the link I just posted.
Thanks for posting that — a great reminder!
Actually: to be honest, I didn’t donate because of the Jimmy Wales thing.
I think the guy is kind of a creep. Had it been posed as “would you like to donate to wiki?” I might have said yes. Since it was “would you like to donate money to wiki because a former wanabe pornographer w/ well known issues thinks it’s a good idea?” – I opted out:)
Unless you never use Wikipedia, your point sounds very hypocritical.
Heh, I totally agree. The second time I saw his creepy mug up there, I went into adblock and blocked all of the banner images. Screw Jimbo. If he wants to run his little private encyclopedia like a vanity project, then he can pay for it.
Funny enough Jimmy’s ad got me to give them money. Two factors came into play, the personal touch from the founder and the fact that Wikipedia is a free service.
Seeing the actual founder “appeal” directly to you felt a lot more personal, almost creating a sense of guilt for the avid Wikipedia user. The other two ads were from the Wikipedia organization instead of a human being. In the words of Seth Godin: “Humans like humans. They hate organizations.”
This type of ads works very well for a free service like Wikipedia, but the same ad would not have performed as well for a for-profit corporations, since people do not relate to CEOs (especially since 2008).
Antoine, I guess you don’t understand that Jimmy’s not really the “founder” of Wikipedia. That’s a lie he cooked up about 4 years after the project was launched by Larry Sanger. Oh, and, Antoine… only 41 cents of each of your dollars will go toward project costs that fulfill the mission you thought you were supporting. Very wasteful.
Nice – I also took a linear approach similar to yours, and added a little bit of interactivity for the user:
Thank you for these challenges! They’re always fun, and they help those with us without the background in the area to learn how to work with data and visualization (finishing up a degree in music composition; it hasn’t exactly prepared me for statistics).
Here’s my attempt – you can click on the wedges to find out more info. I would’ve added more, but spent most of my time focusing on getting the thing to work!
I think the success of the Jimbo banner hinges on how few people realize what a scum ball is Jimmy Wales.
Hi Nathan ! Thanks for sharing the data and inspiring us to create a visualization :)
We decided do something more on the analytics side, trying to find a balance between visual appeal, readability and richness of information.
Here is our blog post http://blog.ffctn.com/our-take-on-winning-wikipedia-fundraiser-bann and you can try the application live here http://ffctn.com/a/ff-wikimedia/.
This is fund raising. Jimmy Wales or any of the founders, this photo worked. Actually the body language in the photo is most compelling. Sure it’s a head shot but visualize that oh so subtle tip of the head. The eyes which suggest, “Don’t beat us down – let us keep providing.” That La Gioconda smile more self-depecating for having to ask. It works. It’s personal.
I couldn’t take it anymore. That ad guilted me into not taking Wikepedia for granted because behind that image is that dare to imagine a world without Wikipedia. The dare to go it alone. (And I’ve donated before)
This is smart advertising.
I felt compelled to donate just to make Jimbo and the banner go away!