Cliff Kuang wrote an interesting article about design at Airbnb called How Airbnb Evolved To Focus On Social Rather Than Searches:

For a couple years, registered Airbnb users have been able to star the properties they browse, and save them to a list. But Gebbia’s team wondered whether just a few tweaks here and there could change engagement, so they changed that star to a heart. To their surprise, engagement went up by a whopping 30%. The star, they realized, was a generic web shorthand and a utilitarian symbol that didn’t carry much weight. The heart, by contrast, was aspirational.

Cliff goes further to explain how Airbnb followed the data clues to understand exactly why the heart performed better, and then used those insights to make some very successful design changes. The data is there, we just have to measure, listen, and respond.

(link via @iamFinch)

One of my favorite sites, The Great Discontent, has a great interview with designer Jason Santa Maria:

The default posture of the Internet is that you put work out and hope that someone connects with it, learns from it, and builds upon it. That isn’t unique to the web community, but it’s one of our community’s greatest traits—everyone shares what they do and we all learn from one another.

From the beginning, whenever I was in a position to tutor or mentor someone, I was always up for it. I want to leave a mark in a way that helps other people to be better and if I have knowledge that can do that, I think I have to share it. By doing so, it sets an example for others to do the same. It pays it forward and helps foster a better community.

I’m not generally a fan of interview posts but this site does them really well, and Jason’s story is inspiring.

Nicholas Carr wrote an excellent, balanced article on the rise of massive open online courses (MOOCs¹) like Coursera and Udacity, and the complex data mining required to make it work. From The Crisis in Higher Education:

The advances in tutoring programs promise to help many college, high-school, and even elementary students master basic concepts. One-on-one instruction has long been known to provide substantial educational benefits, but its high cost has constrained its use, particularly in public schools. It’s likely that if computers are used in place of teachers, many more students will be able to enjoy the benefits of tutoring. According to one recent study of undergraduates taking statistics courses at public universities, the latest of the online tutoring systems seem to produce roughly the same results as face-to-
face instruction.

This is some really in-depth reporting, and it’s not all sunshine and roses. Nicholas went out of his way to seek out and report on legitimate counterarguments to this movement as well.

I’ve long been fascinated by the Dunning–Kruger effect and its distant cousin the Peter Principle. If you haven’t heard of these theories yet, I recommend you don’t read about it at bedtime if you value sleep. This is the kind of thing that keeps you up for days as you try to figure out how it applies to everything you’ve ever done.

Dunning-Kruger basically states that people who are incompetent don’t realise that they’re incompetent, because they lack the competence to figure it out. That’s really scary stuff.

Anyway, in June 2010 Errol Morris conducted an interview with David Dunning, and it’s a fascinating read. Among other things, Dunning gives more background about the research they did, and also goes into detail on the idea of “unknown unknowns”, that scary realm of not knowing what you don’t know. From The Anosognosic’s Dilemma: Something’s Wrong but You’ll Never Know What It Is:

Unknown unknown solutions haunt the mediocre without their knowledge. The average detective does not realize the clues he or she neglects. The mediocre doctor is not aware of the diagnostic possibilities or treatments never considered. The run-of-the-mill lawyer fails to recognize the winning legal argument that is out there. People fail to reach their potential as professionals, lovers, parents and people simply because they are not aware of the possible.

This is a five-part series, and I’ve only read part 1, but I’m really looking forward to digging into the rest of the series. If you have an interest in human behavior, and you’re not scared of freaking yourself out a bit, this is highly recommended reading.

(link via @berkun)

Most successful applications do a good job of onboarding users to teach them how the basics work. After that, good applications also make it easy to learn more advanced features simply through repeated use. You might make a wrong turn once, but if the application corrects your course, you never make that mistake again.

But sometimes there are features that fall between the cracks of onboarding and self-learning. It usually happens when there is some unique behavior in the app that is not only presumed to be commonly known by all users in the community, but is also small enough so that it’s not worth making a big deal out of during new user onboarding.

I recently thought of two such examples that I wanted to share, along with some suggestions for addressing the issue.

Twitter mentions

First, there is the issue of Twitter mentions. I still see people who I know have been on Twitter for years, who don’t know that if they start a tweet with “@”, not all their followers will see it. This information is buried deep in Twitter’s Help section, where I’m guessing very few people venture to. From Types of Tweets and Where They Appear:

Users will see @replies in their Home timeline if they are following both the sender and recipient of the update. Otherwise, they won’t see the @reply unless they visit the sender’s Profile page. 

This is fairly clear, but if you don’t think about this as an issue, you won’t know to ask the question, so it’s not information you’re likely to seek out.

Instagram replies

Second, there is replying to comments in Instagram, which I’m sure trips up quite a few people. If you comment on one of my photos in Instagram, I will get a notification. But if I respond to your comment without including your @username, you won’t get a notification. This is not how it works on Facebook, where you get notified of five comments after the one you posted¹. Instagram does have an easy way to reply to people with their usernames, but it’s a slide gesture I discovered by accident:

So the easiest way to reply to someone is to slide from left to right on their comment, then tap on the arrow. Or you can start the comment with an @, which will then autocomplete the name as you type. But it’s not something they tell you about explicitly. It’s also, again, not information most people will seek out actively, since they’re getting notifications for each comment on their own photos, so why worry?

A proposal

My proposed solution for this type of situation is fairly simple. In the case of features that don’t behave as people expect them to, show a lightbox-type message to explain how it works just one time — the first time they perform the action. For example, the first time a user sends a tweet that starts with an @, show a message to explain who will see it. And the first time a user comments on one of their own photos in Instagram, show a message that explains when people get app notifications.

These are small but important details, especially for social services where understanding exactly what happens when you hit “Post” is essential to the enjoyment of the app.

Related post from the Elezea archive: Best practices for user onboarding on mobile touchscreen applications.

Naomi Canton’s Cell phone culture: How cultural differences affect mobile use is a fascinating article by itself, but the videos and photo slide show really drive home how ubiquitous mobile phones have become all over the world. For example, here are some interviews with cell phone users in Kenya:

Direct link to video on CNN

Mark Helprin offers up some great advice to writers in Skip the Paris Cafés And Get a Good Pen, but it’s advice that works just as well for all creative pursuits:

Your most important tools will be your honesty, labor, courage, practice, luck and utter concentration. […] More valuable than speed or being struck by what you think is lightning (and others usually do not) is concentration. When asked how he managed to come up with the calculus, surely one of the greatest achievements possible for the mortal mind, Newton replied, “I thought of nothing else.”

I love the writing style in this piece. For example, while expanding on his advice not to try to be Hemingway by writing in cafés all over Europe:

Literary skill, much less greatness, cannot be had with a pose, and exhibitionism extorts the price of failure. Also, have pity on the weary Parisians who have wanted only a citron pressé but have been unable to find a café where every single seat is not occupied by an American publicly carrying on a torrid affair with his moleskin.

When I grow up, I want to write like that.

I spent the past two days running usability tests on websites that sell financial products like life cover, funeral policies, and annuities. The target market is lower-income users who access the Internet at least once a day on a desktop at home or work, or on their phones. They are, for the most part, tech literate, and very used to finding their way around the Internet. I wanted to document some of the more general findings while we’re knee-deep in analysis and everything is still fresh.

What follows is a list of interaction design elements that I believe should never, ever be used on a website. They might seem like small issues, but I’ve seen time and again how small things add up, and eventually end with frustrated users who abandon a site altogether. Also, if you’re tempted to think that your users are different and somehow more sophisticated than the ones we tested, please consider the growing digital usability divide.

So, here it is — an incomplete, top-of-mind list of usability sins your website should never commit, based on data gathered through in-person usability testing:

• Don’t use an asterisk (*) to mark required form fields — especially if there is no explanation of what the asterisk means. Most users do not understand this at all. Instead, state that all fields are required unless indicated otherwise, and then mark optional fields with the word (optional). By the way, Luke agrees with me on this one.
• Don’t open links in new browser tabs. Tabbed browsing is for advanced users. If you open a page in a new tab, most users will get lost, start clicking the back button, and then not understand why they can’t get back to where they started. Remember that they’re not focused on the chrome when they click a link, they’re focused on where they’re clicking. So it’s very easy to miss the fact that a new tab has opened.
• Don’t have an FAQ page. Most users don’t know what FAQ stands for, and besides, it’s bad practice to answer questions outside the context people want to ask them in. Figure out where in the process each question in your FAQ might come up, and provide the answer right there within the flow. Don’t expect people to click to a different page to find the information they need.
• Don’t use PDFs at all (unless you’re explicitly stating that it’s a downloadable research paper or something). Many users have no idea what a PDF is, and can’t even tell when they’ve clicked on one. There’s no reason to have your rates/menus/timetables as a PDF as opposed to standard text. This was a recurring theme, but one user in particular clicked on a PDF, didn’t realize it, and continued interacting with it as if they were still on the website.
• Don’t give table rows highlighting mouse-overs if the rows aren’t clickable. This confuses users. Any mouse-over movement gives users a trigger that they can click on the thing. Don’t think they’ll look at the cursor and distinguish between an arrow and a hand — most don’t look past the hover effect.

This is obviously a fairly random list of UI transgressions, but I feel like we talk about the big issues so often that we tend to skim over the smaller ones that can really add up. If you were observing the usability tests we ran this week, you would have felt the same way I did when you saw person after person struggling with the most standard of UI conventions. Let’s just not do these things, for the love of the web and everyone who uses it.