Lessons learned from wikipedia's fundraising banner design | Hacker News Reader

Lessons learned from wikipedia's fundraising banner design

meta.wikimedia.org

94 pointsalexdong37 comments

If you are surprised why the “Personal Appeal” banner is consistently doing so well – well, it looks like this: http://meta.wikimedia.org/w/index.php?title=Special:NoticeTe... and the alternatives look like this: http://meta.wikimedia.org/w/index.php?title=Special:NoticeTe...

I'm amused by the fact that I'm drawn into the picture of Jimmy Wales. It just looks so inspirational; for what reason, I'm not really sure.

the reason? photoshop

(and a decent photographer)

Photoshop is only the means to reach the end goal if you know what that end goal is. What colors, what lighting techniques etc. Give me the source photo and photoshop and I probably couldn't get there in a week, simply because I don't know what elements in that photo are the ones that move people.

Also more important I think are the posture and the look on Jimmy's face which should probably be attributed to good photography.

Personally, I found it the "Jimmy Appeal" banner the most irritating and invasive out of all the different methods they've used. Because of this (negative emotion), it was the only appeal that really affected me. So this morning I was wondering whether Wikipedia folks do test the impact of different methods.

I was also wondering how an attractive looking model would affect the contributions. Certainly worth the experiment.

They should try 'the parked domain girl'.

Yes. Showing a face (or to a lesser degree) demands MUCH more attention that text and simple graphics.

I don't think that was the point being addressed. I think the point was that the personal appeal banner was better designed and much higher contrast (greater use of color, larger font sizes, gradients, etc.), whereas the other banners are lower contrast with smaller fonts and far more ignorable, regardless of the presence of a face.

I know at least a dozen or so of those click-throughs are me trying to hit the little "X" dismiss button in the upper-right corner. Yes I know Wikipedia needs funding, no I can't offer any right now. If I've closed it a half dozen times please don't show it to me for at least a couple page views.

If you're registered Wikipedia user, there's a little option[1] in user preferences that suppress display of fundraising banner.

[1]: http://cl.ly/1J08203O2C0l0R0x2v3V

and if, like the vast majority of wikipedia traffic you're not, there's not.

One more interesting coverage about this design is on quora: http://www.quora.com/Wikipedia/Is-having-Jimmy-Wales-picture... The one with Jimmy on it has 10 times more "conversion" than others. Impressive!

I also like how the fundraising committee involves the community to design the banners. It feels like a genetic algorithm trained/evolved based on human clicks.

Any sufficiently advanced A/B testing setup is indistinguishable from Strong AI -- it just has people embedded in the loop.

Strong AI is overrated: evolution beats intelligent design every time.

It doesn't.

[EDIT: There isn't much more to say. The claim that (esp. natural) evolutionary processes strive for streamlined and optimal solutions is simply false, for various (even a priori) reasons.]

You should point out a counterexample. I guess it would reveal that your definition of optimality is different from "evolutionary fitness", and it shouldn't be.

Note I didn't claim that "evolutionary processes strive for streamlined and optimal solutions". I have no idea how you derived that or why are all these people voting you up and me down (I'd appreciate an explanation btw). I just made a joke that I think contains a kernel of non-trivial truth, although I don't believe it's true if you take it literally as a claim for a universal fact.

Edit: Perhaps I figured out part of the voting - people think I disagree with a popular HNer in an irrelevant way. Amusing.

Just for one thing: evolution is not going to help YOU in this life. Intelligently designed software does.

why the fuck doesn't the banner on the english language site direct to the ENGLISH "personal appeal"?

ip based localization is fucking imbecilic.

What mystified me was all those mentions of {{$sitename}}. Is there an english community-based site somewhere that shares the funds, but not the name, of Wikipedia?

Donations go to Wikimedia, so I'd assume the funding gets split among all their projects[0]. If you take a look at, say, Wikibooks right now, you'll see the same banner there.

[0]: http://en.wikipedia.org/wiki/Wikimedia#Projects

Portuguese Wikipedia: 2 donations, $ 30,00

Because it's hard for Brazilians to donate via PayPal or credit card.

Always remember that donations schemes are not one-size-fits-all.

Or Brazilians just hate Jimmy Wales, or...

It's a bit preemptive to draw a conclusion like that.

Not preemptive, it's slow as hell to link a paypal account to your bank account and few people use credit cards because of the ridiculous interest rates.

I've been through this a lot before a couple of freelancer jobs filled my paypal account.

It's not hard for Brazilians to donate via PayPal or credit card. It's as easy as someone in the USA. I donated today (I probably clicked on the English page, though) via PayPal.

Anybody else bothered by the lack of p-values on these tables? Kudos to the authors for doing some research to see whether there are differences between the performance of the various landing page and banner designs... but the way they're reported their results leaves a lot to be desired.

The the stated differences (in terms of mean donation as well as conversion rate) between banners and landing pages strike me as pretty small (e.g., $26.92 vs. $27.07)- confidence intervals would be helpful here, as I suspect that the differences aren't significant.

Also, is mean really the right metric to be reporting? I suspect that these data aren't normally distributed, or, at the very least, have some outliers on either side of the spectrum (some people who donated hardly anything, plus a few "high rollers"), so it seems to me that the median would be a more informative statistic.

My only concern with these fund-raisers for Wikimedia / Wikipedia are they never publicly disclose their financials. You gotta dig around wikimediafoundation.org for details.

Any fund-raiser for these non-profits, including wikimedia, should at least include links to a balance sheet and profit/loss statement.

Agreed. I'm glad they're at least posting these findings, it's pretty interesting data.

Huh? There's a big difference between 'never publicly disclose their financials' and 'gotta dig around wikimediafoundation.org for details'. It only takes a few clicks and searches to find lots of detail about their past and plans.

Also, as a 501(c)3 nonprofit, Wikimedia's annual '990' tax form from prior years is a public record, available (among other places) at nonprofit directory site GuideStar.org.

They could make the info more prominent -- but what if including a link to deep detail results in fewer donations, whether because it distracts people or somehow convinces them their donation is insignificant?

As long as they're being honest and the info is available somewhere, their fund-raising campaigns should be free from other prominent requirements.

it was a pretty quick google search: http://wikimediafoundation.org/wiki/Financial_reports

Yeah... but if you're going to put up your face and ask for money, it would make people a lot more comfortable if you explained how much of that goes directly in your pocket and what you do to earn it.

It is awesome to see them being so consistent and publishing all of this. I've never seen a non-profit do so in such detail. Some of the comments here (particularly the stats) might be very welcome there, on the talk page or elsewhere. Presumably some at least will do that, and that makes them better able to learn and understand than similarly-sized and staffed organizations would be.

They're just awesome.

I'm curious which versions of each type of banner did the best... (If I missed this in TFA, please let me know.)

To me, the first one I saw, where he is standing off to the right, wearing a black shirt, in what looks like some sort of industrial area, is the most effective.

I know it's Wikipedia, but it's still impressive that they can get 50k donations in an hour.

I'm wondering why only in the italian version of the ad he is called Jimbo Wales... weird.

I thought he looked slightly zombie in this one, it freaked me out a bit :)

http://images.huffingtonpost.com/2010-11-15-wales2.png

totally agree!

Also I browse with chrome at 144% zoom so that picture ended up taking more than half my screen which freaked me out some more.

Reply on news.ycombinator.com