The GiveWell Blog

Request for input

We’re planning to redesign our website in early 2015. We last worked on our website in 2009, and it’s time to refresh it.

Please let us know if you have any suggestions you’d like us to consider. In particular:

  • Is there functionality you wish our website had?
  • Is there information you wish were easier to find?
  • Is there a message you think we should be more actively communicating?

If you have thoughts, please share them either via blog comment or by emailing us at info@givewell.org.

Comments

  • Ben Millwood on December 29, 2014 at 6:48 pm said:

    Since soliciting comments like these has a tendency to overemphasise requests for change, I’m going to remark that I am entirely happy with my interaction with the GW website (which is almost entirely through the blog RSS feed, although I occasionally poke through other bits as well). One unusual point I enjoy is how prominent the Mistakes section is. I really wish more organisations would follow your example in that regard.

  • I suggest highlighting the distinction between givewell and openphilanthropy with a more prominent link to the latter for those more interested in it.

  • I like the GiveWell site; I’d personally hope for a redesign to include some thought into who the site is for, and an expansion of the site to target more people who aren’t like me.

    Most (almost all?) pages on the site are dense text without video, infographics, etc, which results in only attracting people who enjoy and are trained in reading pages of dense text.

  • Adding to what Chris and Ben said, pages for people unlike us would probably be an asset with regard to EA outreach. However, I would like (1) the two kinds of presentation to remain separate, and (2) the spruced-up kind of data to be just as balanced.

    Animal Charity Evaluators is doing point 1 well in my opinion, but some shortcomings in point 2 render their surface-level summaries much less valuable to me than their full reviews. I can imagine that it can be hard to keep pros and cons in proportion when the content is so much more condensed, but maybe some good designer can find a way.

    Then the versioning of the reviews could be handled better. It looks all very manual at this stage. A wiki-like structure with a history built in would be more intuitive to read, easier to maintain, and maybe also easier for search engines to understand.

    Next there are some data on review pages that I perceive as perpendicular to the reviews, for example, whether a review is likely to be outdated or whether a charity is still considered outstanding. These are data that, I feel, should not be part of the review but rather set at some central point and then added to and removed from the review pages automatically.

    Integrating a more feature-rich software like Disqus for comments would also help. Then we’d have threading, we could edit comments, we wouldn’t have to fill in captchas, etc., but that should be a rather small and quick change.

    There are also some blog posts of almost timeless relevance and some that provide updates that are no longer particularly relevant a few months or a year later. It would be handy to have an overview of the first kind (sorry if that already exists somewhere). You could use the category feature for that.

    Oh, the “qualitative” pie chart on the homepage always strikes me as a little silly. ^^

    These are the only things I can think of for the moment. All in all I’m very happy with the sorts of priorities the current design sets (e.g., what Ben said). Those should be as central in the new design as they are now.

  • Aaron Gertler on December 30, 2014 at 5:47 pm said:

    Two changes that I haven’t seen proposed:

    *I wish a page existed such that I could link a friend who hadn’t ever heard of Givewell to that page and say “this is the best introduction”.

    I’ve used “About Givewell”, “Top Charities”, and “Our Process” to do this in the past, but I wish there were something shorter and not quite as technical, perhaps with images or an attractive sitemap. “This is Givewell. We do X, because Y. We place a special focus on Z. Here are some nice quotes/articles about us. Here are some of our most popular articles.”

    *A slightly larger font size/wider spacing. The articles are long, so there’s a trade-off here, but charity reviews immediately seem imposingly dense, and not everyone has a “zoom in” instinct for web text.

    I second the suggestions of Disqus forums and highlighting the GW/OP distinction.

  • Ryan carey on January 1, 2015 at 5:38 pm said:

    I agree with Ben M that the site is good overall. My main item of feedback is that when I link people to GiveWell for the first time, they should have a better experience. So the front page should be prettier with some more pictures, and then when they click through, presumably to the recommendations, the text should be a bit larger, and again there should be some more images.

    Overall, it’s nice that lots of links are available on sidebars but the double-sidebar approach might be a bit dated now.

    I think I agree with Aaron that the font size is a bit too small (like 1pt), which was probably previously necessary because of low res monitors and the two sidebars.

    I agree that Disqus is decent, because of voting and nested comments.

    Well done on reaching a stage where you have moved so much money with such awesome research that site design is once again a plausible bottleneck to your impact!

    Am happy to provide feedback on any designs, you can figure out how to contact me.

  • Ian Turner on January 2, 2015 at 11:22 am said:

    I’d like to add a voice against Disqus. The software is buggy, doesn’t work well on mobile, and has privacy and security implications.
    http://en.wikipedia.org/wiki/Disqus#Criticism_and_privacy_concerns

  • The site is good, but the commenting system could be improved. Why not integrate with Disqus or Facebook? If Disqus has too many problems, I’m sure a suitable service in could be found.

    Also, there should be a policy banning unsolicited requests for money.

  • Telofy on January 5, 2015 at 1:35 am said:

    I didn’t know about some of these criticisms of Disqus unfortunately. Facebook would likely be worse I imagine. IntenseDebate is another one, but I have no experience with it other than as commenter.

    We have stricter data privacy legislation in Germany, but I haven’t been able to find any similar service hosted here. That would be interesting, though. (They’ll surely also provide the service in English.)

  • Neil P. Quinn on January 11, 2015 at 2:53 pm said:

    I wouldn’t recommend prioritizing commenting enhancements. Posts here almost always get fewer than ten comments (the inaugural open thread being an exception at 27), and the current setup seems to handle that load pretty effectively.

    I *would* recommend prioritizing the following, in order of importance:
    (1) A responsive design which adapts to small screens. Currently, the site is almost unusable on smartphones.
    (2) Readability (particularly a larger font size), which is particularly important for dense analytic texts like charity evaluations.
    (3) A streamlined navigation system. The left sidebar has almost exactly the same content as the top bar; I think such duplication adds little value, and actually harms usability since it makes the navigation seem more complex than it is. Apart from that, the current organization seems pretty clear, although there are small opportunities for improvement (e.g. it’s not immediately clear whose “credibility” and “mistakes” the top bar is referring to. “Our credibility” and “our mistakes” would be better.)

    I don’t have any particular suggestions to make about content. I suspect that’s also true for most of the people who will read this post, because anyone involved enough to frequent the blog will have long since answered all their questions about GiveWell. I think what you need is user studies with people who’ve never heard of GiveWell (but fit the general demographics you want to appeal to). It doesn’t need to be formal; all you need to do is round up some people like that (acquaintances or people in a coffeeshop will do fine), sit down beside them with a notepad, and pay attention to what information they struggle to find or understand.

  • Luke Muehlhauser on January 12, 2015 at 7:46 pm said:

    I like the current GiveWell site. The new design will probably be okay, but I figured I’d comment that I personally wouldn’t mind continuing to use the same website design.

  • As a new visitor who discovered the site via The Life You Can Save, I found the website pretty wonky. Definitely dense with text and data. I’m not put off too much by that, and set up recurring donations as a result of what I learned. However, the website could definitely be more visually compelling to elicit interest on an emotional level. Perhaps highlight personal from each of the top charities. I would imagine that might drive more donations from people inclined more toward connection than analytical data.

  • I agree with Aaron and other commenters. This site doesn’t really draw users in–each of its “about us” pages is full of data and long paragraphs, which is not very enticing to someone who isn’t already sold on the charity.

    Even as someone who knows about the charity and wants to donate money, the site can be hard to use. For instance, when I was donating to GiveWell’s recommended charities this year, I had to read the entire blog post to see that you recommended giving different amounts to different charities. When I came back later to find the recommended allocation, it was listed in small print on the right of the “Top Charities” page, making it difficult to find. Worse still, I donated to all of them individually before finding out that you allowed me to donate directly to GiveWell, and you will distribute the funds in the recommended amount. This is much easier for me and for you, and it also allows you to change your mind about which charities are most effective when you get more data and donate to them instead. On your “Top Charities” page, you should have one big “donate” button that sends money to GiveWell in the recommended allocation, and the Top Charities’ donate buttons should be much smaller.

    If you have the time and money, you should seriously consider A/B testing of copy and layouts as well. Don’t get obsessed with it, but some general testing of a few approaches–should we be pithy or verbose? Should we use charts and graphs, images, or text?–will help your website be more engaging and bring in more donations to the charities that need help the most. If you pick one of those changes, you should try adding more pictures to your site. It looks pretty drab and bland. In fact, you’re actually a very inspiring organization, and you’re not expressing that.

  • * Is there functionality you wish our website had?
    No, current functionality is good.

    * Is there information you wish were easier to find?
    Yes, see below.

    * Is there a message you think we should be more actively communicating?
    No, but current message could be communicated more effectively, see below.

    1. The “happy path” for new visitors should be extremely clear and compelling. The current GiveWell site makes an attempt at this (i.e. follow the big orange buttons) but it is let down by:
    a) a weak pie chart on the homepage
    b) four distracting/redundant panels on the homepage
    c) a very weak statement on the top charities page (“These are evidence-backed, thoroughly vetted, underfunded organizations. We discuss the relative strengths and weaknesses of these organizations in this post. We discuss our process for reaching these recommendations below.”) Bleh. You’ve lost their attention by this point.

    2. GiveWell’s content is great, but it’s incredibly ugly. Posts and pages should be *visually appealing*, and instantly digestible at a high level for those that are time poor. I don’t tend to share GiveWell posts on social media because they are so dry and ugly I know they will be dead in the water. For example, compare the layout, styling, phrasing and digestibility of https://www.givewell.org/criteria vs https://80000hours.org/career-guide/. One example: GiveWell posts tend to be self-referential (i.e. use the word “we” a lot), while new visitors following the “happy path” are probably more interested in how the site is relevant to them (i.e. they should use the word “you”)

  • Allen Sussman on January 17, 2015 at 1:36 am said:

    I spent a while on the website before I realized that there was an easy way to give to your top charities according to your recommended allocations. Drawing even more attention to this on the home page (perhaps putting it front and center), making it clear that the allocations will be done for the user, and then including links there to find out more about how these charities and allocations were chosen may help increase giving.

  • Martin Randall on January 19, 2015 at 7:59 pm said:

    Taking the recent Ebola update as an example, I think even your summary blog post has significantly more detail than a small-time donor like me really needs. Rather than leading with “Over the past couple of months, we…”, you could go with “Q: Should I donate to Ebola containment? A: Not any more”. That’s the first and most important thing I need to know.

    The second thing I need to know is what lessons I can draw from this event for future giving. Here, as I attempt to decode your blog post, you seem to be saying that:
    * Disease surveillance and biosecurity are probably great value for money. You’ll have more recommendations on that later.
    * The containment effort was initially underfunded, and this might be a symptom of third world disease outbreaks compared to other natural disasters (which are normally overfunded).

    But that’s my selfish opinion, and other audiences will seek different highlights.

    I completely agree with everyone about graphics and readability and sharability. Every page (other than the detailed writeups, which should probably stay pretty dry) is a potential marketing moment.

  • I suggest setting up a two row link header at the top, with some the most commonly visited pages linked with large font and the less commonly visited ones in smaller font. E.g., “Top Charities,” “Credibility,” and “Donate” might be large and the rest below in smaller font. Spell out more of the subpages that are commonly visited this way, and lose the mouseover.

    I also agree with the suggestion to remove one column and increase space and font size.

Comments are closed.