Talk:Fundraising 2009/Donation buttons upgrade

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search

Please leave comments and questions about the project plan or implementation here. Rand Montoya 21:14, 20 July 2009 (UTC)[reply]

Testing[edit]

  • Were the initial designs tested? If so, where were the results published?
  • No, had initial drafts that morphed to the current versions after reviewing feedback. Rand Montoya 18:25, 28 July 2009 (UTC)[reply]
  • By, "morphed," do you mean "censored based on the opinions of a small group of critics without any regard to any empirical testing?" 75.55.199.5 18:55, 29 July 2009 (UTC)[reply]
  • You will be listened to more carefully if you are civil. -- sj | translate | +
  • Why are different designs being proposed to be tested sequentially when giving patterns are seasonal and periodic?
  • We are currently reviewing this; it is a time/energy vs. technology discussion. In a perfect world, we'd be able to create a perfect test to craft perfect results on each of the buttons. Sadly, I am working within the constraints of our reality, and the currently proposed sequence of testing may be what we go with, however imperfect. Rand Montoya 18:25, 28 July 2009 (UTC)[reply]
  • How much more perfect does the world have to be to make whatever is emitting the HTML for pages include a uniform random number generation selecting the button and including a token in the URL so you can keep track of which one gets the most click throughs? Have you talked to any engineers about this at all? 75.55.199.5 19:00, 29 July 2009 (UTC)[reply]
  • It looks to me as though the designs are now being tested in parallel. -- sj | translate | +
  • Agreed. We have updated the time line and the process to include parallel testing. Thanks to the tech team for making it happen. Rand Montoya 18:23, 14 August 2009 (UTC)[reply]


  • What is the practical limit on the number of designs that can be tested simultaneously?
  • What are the psychological and user experience implications of the command "Donate Now" contrasted with those of a personal appeal?
  • "Profound" not only comes to mind, but it is strongly supported by last season's giving statistics before and after the personal appeal. 75.55.199.5 19:02, 29 July 2009 (UTC)[reply]
  • It would be interesting to do a similar test with different core button text -- have alternatives been proposed? -- sj | translate | +
  • We are open to testing other buttons, locations, and language; however, at this time, we're limiting this only to the donation button. We can certainly test other things in the future. Right now, we're focused on this big change and then moving immediately into the Annual Fundraiser so our time is limited. Rand Montoya 18:23, 14 August 2009 (UTC)[reply]
Are you able to link the Jimbo Button to http://strategy.wikimedia.org/wiki/Call_for_participation/Appeal_letter for testing now? Please do observe all the independent variables you asked for when you asked for the independent variables needing to be tuned to optimize income. 99.191.73.2 00:33, 14 September 2009 (UTC)[reply]


  • What reasons are there against testing the Cartesian product of all the different graphic designs with the different wordings?
    • Are there experimental design techniques which would obtain the same amount of information from a fewer number of experimental tests?
    • Would the number of experimental trials for the full Cartesian product be substantially more difficult to run than running all the graphic designs except one with the same wording and all the wordings except one with the same graphic design? HowDoIUseUnifiedLogin? 19:47, 21 July 2009 (UTC)[reply]
    • See comment above. We'll weigh our resources and our ability and see what we can do. Rand Montoya 18:25, 28 July 2009 (UTC)[reply]
  • Agreed. We have updated the time line and the process to include parallel testing. Thanks to the tech team for making it happen. Rand Montoya 18:23, 14 August 2009 (UTC)[reply]
  • In the first round you asked for comments on the “Message subtext of article text”, most people frowned upon this idea. Am I right to assume that you won’t be using the subtext in this years funddrive? m:Mark W (Mwpnl) ¦ talk 22:04, 23 July 2009 (UTC)[reply]
  • At this time, we have no plans to alter or test any changes to the article subtext. Rand Montoya 18:25, 28 July 2009 (UTC)[reply]
  • If this button is implemented on all Wikimedia projects, will this also include Wikimedia chapter websites? m:Mark W (Mwpnl) ¦ talk 22:04, 23 July 2009 (UTC)[reply]
  • I'm unsure about this, but currently think it's unlikely. I'll review and update on this. Rand Montoya 18:25, 28 July 2009 (UTC)[reply]
  • Update: At this time, this change is designed for Wikimedia Foundation websites. Rand Montoya 18:23, 14 August 2009 (UTC)[reply]
  • Will it be possible for Wikimedia chapters to change the link of the button to their own local donation-website (e.g. all nl-projects have a link to the Wikimedia NL-donation-website). m:Mark W (Mwpnl) ¦ talk 22:04, 23 July 2009 (UTC)[reply]
  • No, and this will be addressed shortly in the upcoming Chapter Fundraising Agreement and via other discussions. Rand Montoya 18:25, 28 July 2009 (UTC)[reply]


Comments[edit]

Parallel testing. Yay!

By the way, would you mind providing a list of the actual donation amounts (perhaps on a separate page) rather than simply a count and dollar total? The donations have a long tail such that low frequency / high value donations can introduce a great deal of random jitter into the averages. To determine the statistical significance of any change one generally ought to look at more robust statistics such as the median and cumulative distribution. Dragons flight 00:13, 12 August 2009 (UTC)[reply]

  • I'll do you one better. When the final data is ready, we'll post all the proper gift data for download. I'll also clean up the chart when I have appropriate data...but do want to maintain a clean, easy-to-read version for those who want a quick look. Rand Montoya 16:17, 12 August 2009 (UTC)[reply]


Level of intrusiveness[edit]

For the future - I'd be interested to compare the effectiveness of a thinner or differently-shaped button that nevertheless pops out of the grey and blue background of the skin, positioned in a variety of places. For instance, below the search box; as an overlay on top of part of the logo; as a diagonal stripe across the upper-right corner of the screen. -- sj | translate | + 05:45, 14 August 2009 (UTC)[reply]

  • At some future time, I'm happy to discuss alternate testing. Right now, we are going to move carefully as this is already a big change. Rand Montoya 18:23, 14 August 2009 (UTC)[reply]

Continual testing[edit]

It's great to see this sort of focused study being done. I'd be interested in seeing a similar test done on a much smaller set of readers, continually. Say 1% of visitors every 3 weeks are identified, split into N test groups [randomly assigned a modified donation skin from a set of N skins] and shown that skin for the next 3 weeks. You could arrange it so that a given IP would only be shown a test skin once in a given year. -- sj | translate | + 05:45, 14 August 2009 (UTC)[reply]


Language[edit]

"Each donation" might work better with "everyone". Where is this being translated? -- sj | translate | + 05:52, 14 August 2009 (UTC)[reply]

Future tests like this might be more accurate if done across a range of language projects, since not only the social norms and ideals but also the default css used can vary from site to site. -- sj | translate | +
  • Absolutely. I would love to make this happen. Someday. Rand Montoya 18:23, 14 August 2009 (UTC)[reply]

Please rethink several presentational aspects.[edit]

I wanted to start translating the core messages but I quickly gave up because:

  • There are several series of ALL CAPS messages. One should know that ALL CAPS must be achieved by CSS, if at all, not using uppercase characters. This needs to be corrected, before useful translations can be made,
  • There are hardly any message that can be translated:
    1. because without explanation many do not make much sense, and
    2. if translated as they are, they appear extremely ridiculous. They're counterproductive in the cultural enviroment, I am translating to. They may be usable in an US-american environemt, but not for us.

--Purodha Blissenbach 23:58, 22 October 2009 (UTC)[reply]

Would you mind discussing this on the page that's devoted to the core messages translation? (Talk:Fundraising 2009/core messages) It's nice if we have all of these discussions in one place so that other translators can benefit from the comments too. Cbrown1023 talk 01:43, 24 October 2009 (UTC)[reply]

Please make more viral[edit]

The current fundraising page does a good job at encouraging individuals who visit the page to donate. But, I think there are simple ways you could do more to encourage people to support the fundraising campaign.

In general, I think that a good rule of thumb for any campaign action, be it a web page or an email is to:

  1. Encourage others to share the action (e.g., email your friends, post to a news aggregator, etc.
  2. Partake in the action.

I'd like to emphasise that this order is important. The first thing you should get people to do is share this action with others. The second thing you should ask them to do is actually partake in the action personally.

Another way to look at this is to pretend that the web page is an individual's personal fundraising page. Right now you have one fundraiser. So, I think the first goal of the page should be to create more fundraisers to go out and seek people to give funds as well as donate themselves if they can. --Joshuagay 04:59, 29 October 2009 (UTC)[reply]

Contributor quotes[edit]

Seriously, I think the best campaign I've seen is from the 2007 Fundraiser that had a random quote from a donor. We shouldn't be telling people in huge letters to donate, we should use the same style of banners as 2007 with anonymous people telling others to donate. Fundraising 2007/comments/en, sincere comments from readers, are a lot better than all this terrible bragging, capital letter, FOREVER stuff. Of the current proposals, use only the Phase 4 set of Fundraising 2009/core messages/en. As an active editor, I actually liked reading what our readers had to say, but with this, I'm just going to hide the banner. Reywas92 19:45, 8 November 2009 (UTC)[reply]

I like the quotes too. Rd232 11:49, 11 November 2009 (UTC)[reply]

Request for removal[edit]

Is there any possibility that a single Wikipedia might request the removal of the banner or at least an alternative, less in-your-face version? We are currently running a writing competition at af.wikipedia, but the notice thereof is hardly noticeable underneath that huge, shouting banner. Anrie 08:19, 11 November 2009 (UTC)[reply]

Hi Anrie. You may be interested in this post. Cheers. --MZMcBride 09:47, 11 November 2009 (UTC)[reply]

Wikimedia forever[edit]

It is just embarassing that the banner says "Wikipedia" instead of "Wikimedia"... I am just stunned how this could pass through.--Kozuch 15:53, 13 November 2009 (UTC)[reply]

I'm sure that the choice is not arbitrary. Also, it appears that banners are only currently being shown on Wikipedias. Dragons flight 17:43, 13 November 2009 (UTC)[reply]