Yes, you're absolutely right! And actually, much more recently we have been been trying to steer our work towards addressing point (2) you raise above.
It became apparent to us, as you say, that most people will find this stuff simply by asking google specific questions e.g. "how many people are there in the world?". And a lot of our existing content will not appear in Google or Siri as the answer, despite us having it in longer articles on the site.
So we've been very recently trying to answer much more explicit questions that people would ask (although we should have been doing this much earlier), and are thinking of reformatting a lot of our existing content in this way too.
Thanks so much for sharing your thoughts - it's really helpful perspective for us.
Thank you both for flagging this. I agree it's not clear at the moment. We are doing some tweaks on the interactive charts and will make sure to include this labelling.
It's funny that you make these suggestions, because they're all things we're grappling with/considering at the moment.
Until now, most of our work has been supply-driven: we write and work on what we think people want to know rather than actually asking them. This has probably been a mistake. We're really changing the way we work to be more demand-driven. This means possibly having some kind of area where people can ask questions, have discussions etc. It's definitely something we're considering.
And you're also right that we need to work on the connectivity between aspects of the site. We're in the middle of trying different designs and formats to try to nail down the best way to do this.
Thanks for sharing your experiences on this. I hadn't heard of Akvopedia before, so it's new to me.
Agree that finding the right solution that sticks: is read by many, consistently kept up-to-date and gets the level so detail correct/unbiased is difficult to do.
We're trying our best to do our version of this work. It's reaching some people, but we can always do more/better. It's really helpful to hear experiences/responses from elsewhere which we can learn what works and doesn't.
As you say, Gapminder is one organization that takes data from published datasets and communicates it to a larger audience. The Gapminder foundation therefore has a similar mission – promoting fact-based world views, informing people about long-term changes in living standards. But Gapminder is not broadly focussed on global change – it does not cover violence, war, poverty, education, environmental change, etc. – and instead more narrowly focuses on health and demography. The areas of my own personal work: environment, food, and energy, for example, are not extensively covered there.
In fact, for those topics that lie outside of its focus, they often rely on us as input (you can find us referenced throughout their book 'Factfulness' for example). And similarly for us: for long-run trends on aspects such as child mortality or fertility rate, we reference their datasets (https://ourworldindata.org/grapher/child-mortality).
So our mission is very similar and we continue to collaborate, but our approach slightly different. Gapminder is narrower in focus and tries to capture a much more general audience. We go for more breadth and depth, and target a less wide (but still large) general audience.
You're absolutely correct that we need to be very careful to ensure bias does not creep into our work.
We take a lot of time to ensure that our work is as objective as possible. Our core activity is to present the data on a given topic clearly. Where possible we try to provide a summary of what the academic research says on this topic: why a given change might have happened, why it happened in a certain way, if we know anything about direct causations. However, causation is particularly difficult: especially when there are so many dimensions interacting (e.g. poverty impacts on fertility, child mortality, education, health outcomes, and their knock-on or cyclical impacts) and a range of interacting external factors (policy decisions, trade liberalisation etc.). Unless there is solid research to support an explanation of change, we don't include it, and stick what has changed rather than why.
Maintaining a very objective, somewhat central position is important for our work. We're happy that both people on the right and left use and trust our work: if we can at least first agree on the facts and data, then we are all in a better position to then discuss why given changes have happened. Media Bias Check (https://mediabiasfactcheck.com/our-world-in-data/) puts us very slightly to the left, but we're very happy that 80% of the user responses rate us 'least biased' i.e. straight down the middle.
In terms of dismissing legitimate concerns about the current state: we think it's really important that acknowledging progress in many areas does not translate into an acceptance of the current state of the world. As we write here (https://ourworldindata.org/much-better-awful-can-be-better) although we have seen improvements in many ways, the state of the world is still unacceptable. And in some cases, things are sliding backwards (e.g. https://ourworldindata.org/homelessness-rise-england). We need to do much more. We hope that in providing a historical outlook on how things have changed provides insights and understanding of how we make the world a better place moving forward.
That's great to hear - we're working hard to make sure our work also reaches many of the key people influencing policy initiatives and decisions.
That's a great suggestion - we hadn't actually thought of that in the context of our work. But it makes sense given we have such an abundance of open-source data.
What Kaggle Kernels do is awesome. It'd be cool to integrate this concept into our work at some point. Thanks for the suggestion!
Absolutely, it would also allow people to have a conversation about the data through various kernels and threads. As another comment said, there was some recent controversy over some of the reports, a Kernel system would enable an environment where people can have a data driven debate about what the data really implies. Thanks for your work and I'm really looking forward to seeing how it evolves!
Edit: When you provide an open platform for people to collaborate, it can be amazing who shows up and what gets done!
Any really honest feedback on the usability of our website would be the best thing for us.
We have lots of content - spanning everything from population growth to plastic pollution, income inequality to cancer rates. This is of course a key part of our work. But this could make it difficult for people to find what they're looking for.
Do you manage to find it easy to navigate? Is there anything you would recommend we do to make it easier for people to find content?
They give you videos with people's out loud thoughts about for example trying to accomplish specific goals with your sites, which should be really useful.
And for a more guerrilla approach you might find this article useful:
Great recommendations. Thanks for the tips. We've found it really useful recently to sit next to someone in person and ask them to navigate. It reveals a lot of unexpected outcomes just by watching them use the site.
But of course, in-person testing can be tricky to scale so an online service is exactly what we're looking for. We'll check them out!
I think the navigation bar is very busy. It shows up over multiple lines, with different sized and shaded buttons. On mobile/tablet, it takes multiple screen lengths. It seems overwhelming and unfocused.
Perhaps because the navigation is overwhelming, you decided to duplicate your dropdown menu for most of the landing page.
I'm not sure why there is such an emphasis on your media coverage and authors at the top of the landing page. Usually that is a tactic to build social proof for customers feel comfortable buying from a new/unknown startup, but as far as I can tell you're not trying to make a sale.
It took me several clicks to get to an article with an interactive graph -- the cool part. More than just the data, you had good writing in catchy, accessible prose.
My suggestions would be
1) to highlight one cool / recent article+graph at the top of the landing page.
2) Simplify your topic organization; you have 16 different high-level topics, with further subtopics. That relegates you to an unwieldy dropdown. I realize you may be using the SDGs, but I'd prefer 5-6 MECE categories like Environment, Health, Economic Development... into which you can organize SDGs. If you had fewer categories you could fit navigation tabs or make use of UI techniques.
3) Overall, decide on what flow you want from the user and design around it. You have a newsletter, donation, even another web domain for SDG tracking. Who are your primary constituents, and what are you trying to help them accomplish? Is it to delight and educate laypersons, maximize small-dollar donations, highlight the academic rigor of the authors' analysis...?
As a reader, I would come back to the site often if it focused on showing me all the coolest data analysis. I might even browse there instead of Hacker News.
Right now, though, it seems closer in my mind to the World Bank's Open Data website, which has an overwhelming amount of great open data, which I would only visit if I were looking to access a raw dataset.
If lay readership is your primary goal, I would suggest looking for design inspiration from media publications like https://fivethirtyeight.com/ or https://www.newyorker.com/ . Even xkcd.com, which has dated design language, achieves the UX of putting the catchy content up front, and providing navigation for people who want more, rather than the other way around.
Thanks so much for this detailed feedback on the website and our design flow. It's really valuable to get an honest critique of how people find using the site.
These questions/issues of navigation, landing page, structure are all things we're grappling with at the moment and have been making small tweaks then reviewing what difference this makes to how visitors explore.
The rationale for highlighting media coverage and academic citations of our work near the top of the landing was to highlight that we're a trusted source for many publications (and actually from media on both the right and left). Credibility and trust are of course important to our work. But it's interesting to hear it wasn't an important factor for you as a visitor.
Thanks again for your thoughts. We're working on all of this at the moment, and will incorporate all of these suggestions into our discussions of how to do this best.
I think the rigor of the data-driven analysis speaks for itself in terms of credibility and trust, and is the unique strength of your initiative to highlight. Your 'about' page is already very strong to demonstrate your organization and affiliations.
It became apparent to us, as you say, that most people will find this stuff simply by asking google specific questions e.g. "how many people are there in the world?". And a lot of our existing content will not appear in Google or Siri as the answer, despite us having it in longer articles on the site.
So we've been very recently trying to answer much more explicit questions that people would ask (although we should have been doing this much earlier), and are thinking of reformatting a lot of our existing content in this way too.
Thanks so much for sharing your thoughts - it's really helpful perspective for us.