I’m one of the lucky ones. I have a great doctor and good health insurance.
A boring generic healthcare company bought my longterm doctor’s group practice a few months ago. First thing they did was screw up the online patient portal, changing it from the poorly designed, barely usable mess I’d learned to navigate to a slightly more polished but somehow blander portal that instantly got hacked. In consequence, they seem to have hired an Internet security firm that advised them to make changes they apparently didn’t understand how to execute. Thus, sign-in was broken for two months. Doctors kept sending patient results to the site, but patients couldn’t access them, and nobody told the doctors. You’d try to explain the problem to a phone receptionist, but if it ever got to the doctor, it was likely phrased as “Another one complaining about the website.”
The site’s makers apparently weren’t informed of the problem for some time, and there was no way to find out who they were to contact them, since there was no contact information available until you signed in, which no one could. Healthcare in America, 2019.
Anyway, they seem to have fixed a couple of the nonfunctioning loops that would prompt you to create a new password and then not recognize that you had done so and prompt you to create a new password and then not recognize that you had done so and prompt you to create a new password and then not recognize that you had done so and…
So today I was able to create a password, almost get 2FA to work, and reorder medication I’d been doing without. Yay!
Living in New York and working in media, I talk to nonprofit organizations a lot. Big or small, they all say the same. No matter how much work they put into their apps and websites, they just don’t get enough new members. No matter how many expensive redesigns they undertake, they still don’t convert. Why is this?
Generally, it’s the same reason any site with a great product doesn’t convert: the organization spends too much time and effort on the pages and sections that matter to the organization, and too little on the interactions that matter to the member. (“Member” is NGOese for customer.)
Of course there are sites that don’t convert because they have a crappy product. Or an inappropriately priced product. Or because their content attracts people who are never going to be their customers, and gets missed by people who might want what they’re selling. Or because their content attracts nobody. Failure has a thousand fathers, and most businesses fail, so the fact that a website doesn’t convert could mean almost anything. (To know what it actually means, you need data, and you need to watch users interact with it.)
But with nonprofit sites, the product is almost always great, and the person visiting is almost always interested. So what goes wrong?
Never mind the user, here’s the About page
What goes wrong is that nonprofit stakeholders are so passionate about their mission—a passion that only deepens, the longer they work there—that they design an experience which reflects their passion for the mission, instead of one which maps to a member’s mental model.
NGOs lavish attention on their About page and mission statement and forget to work on their members’ immediate, transactional needs. And this is true even for those members who are as passionate about the cause, in their own way, as the stakeholders are in theirs. In the wake of a hurricane, a passionate member thinks of your site in hopes of donating food or giving blood. But nothing on the site calls out to that member and addresses her needs. All she sees are menus, headlines, and buttons trying to lead her to what matters to the organization—namely, the things it says about itself.
How to satisfy the user and convert at the same time
First, decide what one single action you, as the organization, want the user to perform. Should they sign up for your mailing list? Make a donation? Keep it singular, and make it simple. One form field to fill out is better than two; two is better than four.
Next, put yourself in the member’s shoes. What does that member wish to achieve on your website? Have you created transactions and content that allow her to do what she came to do? Have you designed and written menus, links, and headlines that help her find the content that matters to her? Forget the organization, for now. Pretend the only thing that matters is what the user wants. (Because, ultimately, it is.)
Do these things, and weave your singular, simple conversion opportunity into each screen sequence with which your user interacts. To optimize your chance of success, place the conversion opportunity at the very point where the user successfully finishes transacting the business that mattered to her. Not before (where it is only a distraction). Not in another part of the site (which she has no interest in visiting). She’s a lot likelier to sign up for your mailing list after you’ve helped her donate food to her neighbors than she is to sign up in an unsolicited popup window.
Thank you, Captain Obvious
All the above suggestions are obvious common sense, and have been known since transactional web design was in its infancy in the 1990s. And yet, because of organizational dynamics, internal politics, and our getting so close to our own material that our eyes go out of focus, we forget these simple ideas more often than we use them—and fail when success is so easy, and so close to hand.
THIS year’s Poynter Digital Newspaper Design Challenge was an attempt by several designers and pundits, working and thinking in parallel, to save real news via design. In Part 1 of my report from Poynter, I discussed the questions driving the challenge, and talked about the design work done in response to it by my colleagues Kat Downs Mulder, Mike Swartz, Lucie Lacava, and Jared Cocken. Here in Part 2, I’ll discuss my own work and the approach we took at my studio. But we begin with a quick look back at the past designs that brought us to this point:
Experiment 1: The Deck
During the past decade and a half, as both a publication designer and a publisher, I watched in horror as our publications became reader-hostile minefields of intrusive ads, overlays, and popups. The first thing I tried to do about this (besides removing the web equivalent of chart junk from my magazine) was to offer an alternative approach to advertising via The Deck, an ad network I cofounded with Jim Coudal of coudal.com and Jason Fried of Basecamp (formerly 37signals). The Deck permitted only one appropriately targeted ad per each page of content viewed. As primary instigator Jim Coudal put it:
A buy in The Deck reaches the creative community on the web in an uncluttered, controlled environment, far more valuable than a standard banner or a single text ad among dozens of others.
Jim, Jason, and I hoped that our cost-per-influence model would replace the CPM race to the bottom, and that our quasi-religious use of whitespace would be widely imitated by the smartest publications online.
But that didn’t happen. Advertising just got more intrusive. The Deck succeeded as a small business supporting a network of interesting small publications, but not at all as a primary influencer on the direction taken by advertising that supports web content.
Experiment 2: Readability
Then about seven years ago, my friend Rich Ziade and his engineers created Readability, an app that sat between you and the ugly site you were trying to read, the way screen readers sit between visual websites and blind web users. Readability grabbed an article page’s primary content, removed the junk, and replaced the cluttered and illegible layout with a clean, readable page inspired by the clarity of iBooks and Kindle, which were just taking off at the time.
Rich released Readability 1.0 as open source; Apple immediately absorbed it into the Safari browser, where it continues to provide Safari’s built-in “reader” mode. (Safari’s “reader” mode was Apple’s first step in decluttering the web and returning it to the people who use it; “content blocking” would be the second step.)
Moreover, Readability 2.0, released by Rich’s then-company Arc90 the following year, added automatic payment for content creators slash publishers, as I explained at the time to anyone who would listen. Had Readability been allowed to continue the experiment, content monetization might have been less of a problem than it is today, and publication brands (the notion that it matters who publishes what we read) would be in exactly the same pickle they’re in anyway—except that readers would get their news in Readability’s attractive and customizable format, instead of from Apple News, Facebook, and the like.
I used to go around the world on lecture tours, warning my fellow designers that if we didn’t figure out how to declutter and compellingly brand sites, apps like Readability would do it for us. I still go around on lecture tours, but I’ve moved on to other issues. As for Readability, it was killed by a digital lynch mob after one powerful blogger, misunderstanding the motivation behind it, issued the digerati equivalent of a fatw?. But that’s another story.
Experiment 3: Big Type Revolution
In 2012, inspired by Readability and frustrated by the industry’s determination to make ever less legible, ever more cluttered sites full of tracking and popups and everything except what readers need, I bet big on large type:
This redesign is a response to ebooks, to web type, to mobile, and to wonderful applications like Instapaper and Readability that address the problem of most websites’ pointlessly cluttered interfaces and content-hostile text layouts by actually removing the designer from the equation. (That’s not all these apps do, but it’s one benefit of using them, and it indicates how pathetic much of our web design is when our visitors increasingly turn to third party applications simply to read our sites’ content. It also suggests that those who don’t design for readers might soon not be designing for anyone.)
Writing in Forbes, Anthony Wing Kosner saw the future in my initially crude experiment:
If you want to know where the web is going, one clue is to look at the personal sites of top-tier web designers. And one trend that just bubbled to the surface is large body type—the kind you don’t have to command-plus to read.
Jeffrey Zeldman…made a particularly strong point about it in his “Web Design Manifesto 2012,” that he published yesterday.
Not to brag (okay, too late), but he wasn’t wrong. It was the future.
(Also, I’m fairly sure I wasn’t the only designer at the time who reacted against tiny type and cluttered anti-user layouts by stripping pages down to only their most necessary elements, and boosting the type size to enforce a more relaxed reading posture. The idea was in the air.)
An uncluttered page focused on the reading experience (reminder: big type and plenty of whitespace) is now the default at several leading news publications. But many smaller publications, struggling just to survive, have not kept up. And so we have a perfect crisis:
Publications that do not encourage reading, loyalty, or repeat visits are struggling to survive at the very moment real news is under attack from an authoritarian president. What to do?
My response to the Poynter Design Challenge
There are many ways to respond to an existential crisis like the one facing most news publications. You can rethink the relationship between reader and publication. Rethink the job of the publication. Make news work more like a lifestyle app. Make it more immersive. My colleagues followed those paths out brilliantly (as described in Part 1).
But I went for the low-hanging fruit. The thing any publisher, no matter how cash-strapped, could do. The thing I had seen working since I started yelling about big type in 2012. I went for a clean, uncluttered, authoritative, branded page. Authoritative because this isn’t fake news. Branded because the source matters.
The easiest, fastest, most readily attainable path to clean, uncluttered, authoritative, branded design is through typography.
Any publication can be readable
Any newspaper, however poor, can afford better typography. Any newspaper with a designer on staff can attain it, if the paper stops treating design as a lackey of marketing or editorial or advertising, and sets designers free to create great reading experiences.
In my work, which is still underway (and will continue for some time), I focused on creating what I call “reader” layouts (and probably other designers call them that too; but I just don’t know). Layouts that are branded, authoritative, clean, uncluttered, and easy to read.
I played with type hierarchies and created simple style guides. Most of my little pages began as Typecast templates that I customized. And then Noël Jackson from my studio cleaned up the HTML and CSS to make it more portable. We put the stuff up on GitHub for whoever wants to play with it.
These are only starting points. Any designer can create these kinds of layouts. There’s nothing special about what I did. You can do the same for your paper. (And if you can’t, you can hire us.)
The work I share here is the start of a project that will continue at our studio for a long time. #realnews for the win!
CAN design create a better user experience that engages readers and drives revenue? Can it fight fake news and help save real journalism at a time when news organizations large and small are underfinanced and under attack?
These questions drove the Poynter Design Challenge, “a project to create new visual models for digital news publications” sponsored by William R. Hearst III, hosted by the Poynter Institute, and directed by publication designer Roger Black.
The challenge began October 17–18 in New York, with five pundits and five designers, of whom I was honored to be one, workshopping a project brief during a two-day conference event at the Columbia Journalism School. (You can watch videos of all these sessions courtesy of Fora.tv.)
The next phase took place yesterday in St. Petersburg, Florida, as the four other designers and I presented our work to a live audience. In this short piece, I’ll talk about the designs my colleagues presented; in the next, I’ll discuss my own.
Reconnecting with the people: the challenge for digital news
Roger Black described the difficulties facing digital news publications:
The challenge is serious. Fake news crowds real news. Numbers no longer add up for publishers. Readers jump from site to site without knowing where they are, or staying for long. You can see the brief for this project here.
Can design help? Well, as a I designer, I think it can. I mean, the design of most news pages is not what you’d call attractive. But the solutions proposed at Poynter will be much more strategic than cosmetic. And they’re strategies that can be combined.
—Five design answers that add up, Roger Black, January 20, 2017
“A news publication might think a bit more like Fitbit”
Between us, we designers had about a century of experience designing digital publications—internally, as consultants, or both. This means that, even though an open “design challenge” brief necessarily omits an unknown number of the specific requirements any actual publication design assignment would include, all of us were aware of, and to some degree addressed, typical news publication requirements not included in our brief.
Kat Downs Mulder, Graphics Director at The Washington Post, shared a prototype for a big-brand news site. Kat had just given birth to a healthy baby boy (congratulations!), so her work was presented by two of her colleagues from The Post. Kat did not design with the avid, committed news reader in mind (since those folks are not the problem for most publications). Instead, she pondered how to engage the typically fragmented attention of today’s distracted and passive news reader:
“A big-brand news site [should be] aware that people have a lot more to do in their lives than read the news,” Kat posited. Thus, “A news publication might think a bit more like Fitbit. That is, it should make you feel like it’s working for you. A reader should say, ‘I’m reading everything I need to know.’”
Keep that dopamine pumping
Kat presented a multi-paned prototype. The wider pane on the right contained news content; the narrower pane at left was navigation. As I’ve just described it, this isn’t much different from the current Post website, but Kat’s prototype was very different, because it prized reader control over editorial director control; kept track of what you read; encouraged extra reading the way Fitbit encourages extra steps, and rewarded it the same way Fitbit does, with an accumulation of points that give the reader dopamine hits and create the perception that the “news app” is working for her—as a rewarding part of her busy lifestyle.
An Operating System for your city
Mike Swartz, Partner at Upstatement, a design and engineering studio in Boston, took on the challenge to smaller publications (such as his original hometown paper, the Pittsburgh Post-Gazette) which lack the resources of a Washington Post or New York Times.
Mike’s presentation, “information OS for a city: redefining the opportunity for local media,” turned the journalistic prowess of a good local paper into a superpower, connecting readers to their city the way the “terrible towel” stunt concocted in desperation by radio announcer Myron Cope in 1975 reconnected Pittsburghers to their hometown football team, and helped the Steelers win Super Bowl X over the Dallas Cowboys.
There’s a potential for an operation like the [Post Gazette] to rebrand itself as more of an “informational operating system” for its city. With different types of products that are focused and useful and not necessarily bundled into a traditional news format, we can create more enjoyable experiences and more useful products readers will love.
Building reader interest and finding a way to pay for it all
Where the rest of us avoided the elephant in the room, in her design Lucie Lacava, president of Lacava Design Inc., boldly confronted the challenges of advertising and monetization. Algorithm-driven advertising frustrates users, who, in desperation, block it. Choked for income as a result, publications and advertisers create more and more intrusive forms of unwanted advertising. Nobody wins.
And while subscription models have worked, at least partly, for some of the very top news publications, such models are not likely to help most news publications in the near term.
Digital publication as digital application
Lucie’s design addressed these challenges by recasting the news as a hyper-customized application targeted at younger users, who get to choose news streams and ads that are relevant to them. “The elusive millennial” was Lucie’s target. I cannot do her idea justice with a couple of paragraphs and a single screen shot.
Affordable, immersive VR is here
Jared Cocken, brand and product designer for hire and co-founder of STYLSH.co., approached the “attention war” by showing how any size publication could create “video or VR driven stories that enrich a user’s understanding of the world around them.”
Because VR video is immersive, it holds viewer attention. Because it is reality-based, it fights fake news. (It’s hard to call bullshit on a scene you can explore from any angle.) VR also, potentially, builds compassion. It’s one thing to read about conditions in a Syrian refugee camp, another to visually experience them in VR.
Until now VR and video have been cost-prohibitive, but, working (and co-presenting) with VR startup founder Anna Rose and Hollywood producer/actor Banks Boutté, Jared showed how even woefully under-financed newsrooms can use newly designed, super-affordable tools to create “video or VR-driven stories that enrich a user’s understanding of the world around them.”
Blogging about a conference is like tweeting about a sexual experience. You had to be there. I wanted to record and share the outlines of what my fellow designers presented, but these few paragraphs should in no way be considered authentically representative of the deep thinking and work that went into every presentation.
You may see holes in some of the arguments presented here. In some cases, I might agree with you—some ideas, while dazzlingly creative, did not seem to me like the right way to save news. But in most cases, if an idea seems wrong, blame my telling. If you had been there and heard and seen everything, the value of the proposal would have far more apparent than it can be here.
I love that each of us took on a quite different aspect of the problem, and addressed it using very different tools. I’ll be back soon with a short write-up of the design approach I took. Meanwhile, I want to thank all the pundits, designers, and attendees in New York and St. Petersburg—and the Poynter Institute, Roger Black, and William R. Hearst III for making it all possible.
IN a world where newspapers are dying and half the public believes fake news, what online news experiences need is design that is branded, authoritative, and above all, readable:
Branded, because we need to convert the current hummingbird model (where readers flit from flower to flower) back to the idea that your news source matters—and that it is worth your time to return to a source you trust.
Brand helps the social-media-driven seeker notice that they’re returning time and again to a certain resource, facilitating a mental model shift back toward destination web browsing. When every site looks the same, it’s easy to see all content as equal—all spun from the same amorphous mass. A strong brand, which is individual to the given newspaper, can cut through that amorphousness, which is the first step in building (or rebuilding) loyalty.
Authoritative, because combating fake news means visually cueing the reliability of a particular source—one staffed by real journalists and editors, with real sources in real countries. In the coming years this will be more important than ever.
Readable, because an informed public needs to grasp stories that can’t always be reduced to headlines or sound bytes. Readability means even longer articles actually get read, sometimes even all the way through. Readability requires a combination of typeface, type size, leading, measure, hierarchy, contrast, etc.—as well as the introduction of visual information, both to break up the flow of text, and to further illuminate what is being said.
Related news keeps readers reading
Additionally, this branded, authoritative, readable content needs to become (to use an ancient word) sticky: through a combination of editing and algorithms, related content must be presented at the appropriate time in the reading experience, to engage the visitor in continued reading.
Currently two publications—nytimes.com and medium.com—achieve all these goals better than any other publications on the web. One is the newspaper of record; the other is a vehicle for anyone’s content. Yet both really do the job all newspapers will need to do to survive—and to help the Republic survive these next years. I particularly admire the way both publications surface related content in a way that practically demands additional reading.
Design won’t solve all the problems facing newspapers, but it will help. And unlike more “immersive” approaches such as WebVR, original full-screen imagery, and original embedded video, the basics of solid, readable design should not be out of budgetary reach for even the most cash-strapped news publisher—budget being a problem for any business at any time, but especially for newspapers now.
In my studio, we’ve been pondering these problems for content sites of all types (not only newspapers). At the Poynter Digital Design Challenge next month, I hope to share designs that nudge the conversation along just a bit further.
“You don’t get to decide which platform or device your customers use to access your content: they do.”—Karen McGrane, Content Strategy for Mobile
“When a person tells you that you hurt them, you don’t get to decide that you didn’t.”—Louis C.K.
“Discomfort with others’ burdens has no place in good design.”—Mica McPheeters
“Historically, teams simply have not been trained to imagine their users as different from themselves—not really, not in any sort of deep and empathetic way.”—Sara Wachter-Boettcher
“USER CUSTOMIZATION” on the web hearkens back to the deluded old days of portals, when companies imagined you’d start your daily “net browsing” session by “logging on” to their website’s homepage. Customization was among the chief (largely imaginary) inducements for you to return to their “start” page and not others.
The thought was that changing the fonts and color scheme would make their page feel more like your home. After all, Windows 3.1 users seemed to enjoy switching their home computers to “Black Leather Jacket” or other personalized settings—if only as an escape from the computer environment at work, where their bosses enforced a rigid conformist look and feel, and dictated which software and fonts were allowed on your workstation. Surely, the thinking went, pioneering web explorers would demand custom accommodations as plush as those found in the best-selling operating system.
MySpace … and beyond!
This fetish for pointless customization—customization for its own sake—persisted through the MySpace era, where it actually made sense as an early mass offering of page owner personal branding. Its descendants are the WordPress, Tumblr, and Squarespace themes that create a professional appearance for the websites of individuals and small businesses. This is a positive (and inevitable) evolution, and a perfect denouement for the impulse that began life as “user customization.”
But, except on a few quirky personal sites like Jeremy Keith’s adactio.com, where sidebar customization widgets live on as a winking look back to the early days of personal content on the web, user customization for its own sake has long been out of favor—because experience, referrer logs, and testing have long shown that visitors don’t bother with it.
Perhaps that’s because people don’t really visit websites any more. They drop in quickly on a page found by search or referred by social media, scan quickly and incompletely, and leave, mostly never to return.
When you use Google, Bing, or Duck Duck Go to find out what a knocking sound in your radiator or a pang in your gulliver might mean, you scan for the information you sought, find it (if you’re lucky), and leave. The notion that most sites could get you to come back by offering you the ability to change fonts or colors is self-evidently absurd. Why bother?
Readability and font customization
Ah, but there’s another kind of user customization that I’m hoping and betting will make a comeback: a subtle, inclusive sort of customization that doesn’t exist for its own sake, but rather to serve.
Our glowing, high-density screens are great for watching Westworld, but a bit too bright and backlit for prolonged reading compared to the paper they’re intended to replace. But screens have one advantage over printed books (besides storage and portability): namely, they offer accessibility features a printed book never could.
I once received an architecture book written by an important scholar, but I was never able to read it, because the layout was terrible: the type was too small, the leading too tight, and (most of all) the measure far too wide to be readable. If an ebook version had been available, I’d have purchased it; but this was before the mass market availability of ebooks, and the tome is now out of print. I own it, but I shall never be able to read it.
It wouldn’t be a problem with an ebook, because all ebooks offer readers the ability to alter the contrast and the basic theme (white text on black, black text on white, dark text on a light background); all ebooks offer the ability to adjust font size; and most include the ability to change fonts. Why do Kindle and iBooks offer this flexibility? Because it helps readers who might otherwise not be able to read the text comfortably—or at all. This isn’t customization for its own sake. It’s customization for the sake of inclusion.
The grey lady and user customization
Now notice who else provides some of this same inclusive customization function: the mighty New York Times.
People in our industry tend to repeat things they’ve heard as if they are eternal verities—when the real truth is that each digital experience is different, each person who engages with it is different, and each device used to access each experience brings its own strengths and limitations.
A font size widget may smell like the pointless old-fashioned “user customization” to be found on half the unvisited sites in the Wayback Machine, but it is the very opposite of such stuff. Even mighty responsive design benefits from offering a choice of font sizes—because there are just too many complications between too many screen sizes and device features and too many pairs of eyes to ensure that even the best designer can provide a readable experience for everyone without adding a simple text size widget.
Most of the sites we’ve designed in the past few years have not had a text size widget, but I believe this was due to our privileged assumptions and biases, and not to the reality of the needs of those we serve. Going forward on client projects at studio.zeldman, and in my publications like A List Apart, I hope to correct this—and I hope you will think about it, too.
AT HOME, sick with a cold and bored, my daughter buys a single packet of “My School Dance” in a freemium iTunes game. The manufacturer charges her (well, charges me) for ten packets. This same “accidental” 10x overcharge happens across three different games by the same manufacturer in the span of about an hour.
American Express notifies me of the spurious charges, but won’t let me dispute them until they are “posted.” I spend half an hour on the phone with a very nice gentleman at Amex learning this. Why would Amex notify customers about a charge days before they can do anything to resolve it? I don’t know. And I don’t ask the gentleman on the phone. His job is hard enough.
A few days pass. Amex “posts” the false charges and emails me with a link to resolve the problem on Amex’s “dispute a charge” web service.
Amex’s “dispute a charge” web service “encounters an error” when I try to use it to resolve the problem.
This happens every time I try. I try for three days.
So I call Amex, but I can’t resolve the problem because I don’t have the card in my wallet.
So I head to iTunes, where I should have gone in the first place, and click through two or three generations of iTunes “Report a Problem” interfaces: visually different generations of iTunes software, with different user paths, all still being served by Apple. Generations of iTunes software that, when they fail, link to other generations of iTunes software, which also fail.
I click and click my way through five years of iTunes interfaces.
Finally I find an iTunes page where I can manually “Report a problem” for each of the 27 false charges. (Three of the charges, remember, were legitimate. I’m willing to pay for the three items my daughter intended to buy. But not 30.)
If one software product overcharges your kid by a multiple of 10, that could be a software bug. When three products from the same manufacturer all do it, that’s not a bug, it’s a deliberate attempt to defraud families, by overcharging on purpose and hiding behind the opacity of iTunes’s purchase reporting. Simply put, the manufacturer is dishonest, and figures iTunes’s support section is impenetrable enough that you’ll eventually give up trying to get a refund.
But they didn’t count on my tenacity. I’m the Indiana Jones of this motherfucker. I have studied maps and bribed natives and found my way to the hidden iTunes refund page that actually, sometimes, works.
On this page, I inform Apple of the fraud 27 times, in 27 different boxes. Each time, after reporting, I click a blue button, which generally returns an error message that iTunes was unable to process my request. So I enter the data and click the button again. It’s only 27 boxes of shit. I’ve got all the time in the world.
The page tells me that only two refunds went through. Every other request ends with an error message saying iTunes could not process my request, and encouraging me to try again later.
Instead, I leave the page open, and, about ten minutes later, I manually reload it. When I do so, the display updates—I guess this generation of iTunes software preceded “Ajax”—and I learn that most of my refunds have gone through.
So the software actually works about 33% of the time, even though it indicates that it only works 5% of the time. Remember that wait-ten-minutes-then-randomly-reload-to-see-if-anything-changed trick. It’s the sign of excellently designed consumer software.
I’ve put over two hours of my time into this. Going on billable hours, I’ve probably lost money, even if I get all my overcharges refunded. But there’s a principle here. Several principles, actually. Tricking kids is wrong. Stealing is wrong. Building a beautiful front-end but neglecting customer service is wrong. Mainly, I’ve just had enough of 2016’s bullshit.
12 LESSONS from An Event Apart San Francisco – ? 3: Derek Featherstone was the 10th speaker at An Event Apart San Francisco, which ended Wednesday. His session, Extreme Design, showed how creating great experiences for people with disabilities results in better designs for everyone.
Focusing relentlessly on accessibility helps us think of extreme scenarios and ask questions like “how can we make this work eyes free?” and “how can we make this work with the least amount of typing?” Most importantly, it leads to deeper design thinking that solves problems for everyone who uses our sites and products.
A Map For The Blind
One of my favorite examples from Derek’s presentation had to do with a map. A Canadian city was expanding geographically to encompass some of the surrounding suburbs. The city’s website was charged with letting all citizens know about the change. The web team did what you or I would probably do: they created a map that clearly showed the old and new city limits.
Unfortunately, this visual map was by definition inaccessible to blind citizens, so the city brought in Derek and his colleagues to design an equivalent experience for the unsighted. Derek’s team and the web team pondered typical solutions—such as laborious written descriptions of the city’s shifting geographic borders. But these were not user-friendly, nor did they get to the heart of the problem.
Maybe creating a verbal equivalent of a visual map wasn’t the answer. Derek’s team kept digging. Why was the map created in the first place, they asked. What was the point of it? What were users supposed to take away from it?
It turned out, people wanted to know if their street fell within the new city boundaries because, if it did, then their taxes were going to go up.
Solving for a map wasn’t the point at all. Allowing people to find out if their home address fell inside the new city limits was the point.
A simple data entry form accomplished the task, and was by definition accessible to all users. It was also a much quicker way even for sighted user to get the information they wanted. By solving for an extreme case—people who can’t see this map—the web teams were able to create a design that worked better for everyone.
Tomorrow I’ll be back with another top takeaway from another AEA San Francisco 2016 speaker. The next AEA event, An Event Apart St. Louis, takes place January 30-February 1, 2017.
Gerry shared example after example of cases where most of this stuff didn’t matter at all to the person using the site or service, and drew the commonsense—but too rare in the corporate world—conclusion that if we spend our time making stuff that matters to our organization instead of stuff that matters to our customer, we will lose our customer. (“Nobody reads your annual report.”)
One of my favorite takeaways from Gerry’s session was about performance, but not in the way you probably think. Gerry pointed out that, in organizations, we are always measuring our own performance: how quickly did we turn that project around? Did we launch on time? Instead of dressing up our navel gazing with analytics that are about our tasks, we should measure our customers’ speed. How quickly do our sites and products help our customers achieve their goals? How can we identify and remove additional obstacles to completion, so our customers achieve their goals faster and faster?
We need to manage speed on the page, not just the speed of the page load. Manage the customer’s time on task. We won’t become customer-centric until we change our metrics—focusing on customers’ time to complete tasks, not on internal speed, and not just on the mechanical speed of page load—although page load speed (and perceived page load speed) are also terribly important, of course, and are part of improving the customer’s time to complete their task.
“If you solve the customer’s problem, they’ll solve your problem.” When you understand your customer’s top task, and focus relentlessly on helping them achieve it, you build a relationship that works for organization and customer alike.
Tomorrow I’ll be back with another top takeaway from another AEA San Francisco 2016 speaker. The next AEA event, An Event Apart St. Louis, takes place January 30-February 1, 2017.
MESMERIZED as we have been by the spectacle of the flaming garbage scow of U.S. election news, it would have been easy to miss this other narrative. But in the past few days, just as Google, AT&T, and Time-Warner were poised to turn the phrase “online privacy” into a George Carlin punchline, in marched an unlikely hero to stop them: the American Federal Government. Who have just…
approved broad new privacy rules on Thursday that prevent companies like AT&T and Comcast from collecting and giving out digital information about individuals — such as the websites they visited and the apps they used — in a move that creates landmark protections for internet users.
Given the increasingly deep bonds between corporate overlords and elected officials, this strong assertion of citizens’ right to privacy comes as something of a surprise. It’s especially startling given the way things had been going.
On Friday, Oct. 21, shortly before a massive DDOS attack took out most U.S. websites (but that’s another story), ProPublica reported that Google had quietly demolished its longstanding wall between anonymous online ad tracking and user’s names. I quote ProPublica’s reporting at length because the details matter:
When Google bought the advertising network DoubleClick in 2007, Google founder Sergey Brin said that privacy would be the company’s “number one priority when we contemplate new kinds of advertising products.”
And, for nearly a decade, Google did in fact keep DoubleClick’s massive database of web-browsing records separate by default from the names and other personally identifiable information Google has collected from Gmail and its other login accounts.
But this summer, Google quietly erased that last privacy line in the sand – literally crossing out the lines in its privacy policy that promised to keep the two pots of data separate by default. In its place, Google substituted new language that says browsing habits “may be” combined with what the company learns from the use Gmail and other tools.
The change is enabled by default for new Google accounts. Existing users were prompted to opt-in to the change this summer.
The practical result of the change is that the DoubleClick ads that follow people around on the web may now be customized to them based on your name and other information Google knows about you. It also means that Google could now, if it wished to, build a complete portrait of a user by name, based on everything they write in email, every website they visit and the searches they conduct.
The move is a sea change for Google and a further blow to the online ad industry’s longstanding contention that web tracking is mostly anonymous.
Google has long portrayed itself as one of the good guys, and in many ways it continues to be that. I can’t think of any other insanely powerful mega-corporation that works so hard to advocate web accessibility and performance—although one of its recipes for improved web performance, making up a whole new proprietary markup language and then using its search engine dominance to favor sites that use that language and, of necessity, host their content on Google servers over sites that use standard HTML and host their own content, is hardly a white hat move. But that, too, is another story.
On privacy, certainly, Google had shown ethics and restraint. Which is why their apparent decision to say, “f–– it, everyone else is doing it, let’s stop anonymizing the data we share” came as such an unpleasant shock. And that sense of shock does not even take into account how many hundreds of millions of humans were slated to lose their privacy thanks to Google’s decision. Or just how momentous this change of heart is, given Google’s control and knowledge of our searches, our browsing history, and the contents and correspondents of our email.
Minority Report
Scant days after ProPublica broke the Google story, as a highlight of the proposed merger of AT&T and Time-Warner, came the delightful scenario of TV commercials customized just for you, based on combined knowledge of your web using and TV viewing habits. And while some humans might see it as creepy or even dangerous that the TV they’re watching with their family knows what they were up to on the internet last night, from an advertiser’s point of view the idea made $en$e:
Advertisers want … to combine the data intensity of internet advertising with the clear value and ability to change peoples’ perceptions that you get with a television ad. If you believe in a future where the very, very fine targeting of households or individuals with specific messaging makes economic sense to do at scale, what this merger does is enable that by making more audience available to target in that way.
Into this impending privacy hellscape marched the U.S. Government:
Federal officials approved broad new privacy rules on Thursday that prevent companies like AT&T and Comcast from collecting and giving out digital information about individuals — such as the websites they visited and the apps they used — in a move that creates landmark protections for internet users. …
The new rules require broadband providers to obtain permission from subscribers to gather and give out data on their web browsing, app use, location and financial information. Currently, broadband providers can track users unless those individuals tell them to stop.
The passage of the rules deal a blow to telecommunications and cable companies like AT&T and Comcast, which rely on such user data to serve sophisticated targeted advertising. The fallout may affect AT&T’s $85.4 billion bid for Time Warner, which was announced last week, because one of the stated ambitions of the blockbuster deal was to combine resources to move more forcefully into targeted advertising.
The consequences of these new rules—exactly how advertising will change and networks will comply, the effect on these businesses and those that depend on them (i.e. newspapers), how Google in particular will be effected, who will cheat, who will counter-sue the government, and so on—remain to be seen. But, for the moment, we’re about to have a bit more online privacy and anonymity, not less. At least, more online privacy from advertisers. The government, one assumes, will continue to monitor every little thing we do online.
SINCE AT LEAST 2010, subscribers to Spotify’s paid music service have asked the company to include the ability to sort playlists alphabetically in the desktop player. It’s the sort of drop-dead obvious feature that should have been built into the player while it was still in alpha. Yet, after six years of requests by paying customers, the feature still does not exist. Many good people work at Spotify and take pride in working to create the best possible music service. But the management in charge of feature requests does not seem to care about or respect customers.
Spotify subscribers organize their music in playlists. Any serious music listener will soon have dozens, if not hundreds, of playlists. They appear in the sidebar in reverse chronological order of the date of their creation. From a programmatic standpoint, the order is random. The inability to sort playlists alphabetically soon makes listening to one’s entire collection problematic. You ignore most of your playlists because you can’t find them, and waste time recreating existing playlists because you’ve forgotten they exist—or can’t find them.
For years, Spotify users have taken to the company’s message boards to request that this basic, rudimentary, obviously necessary feature be added. And for years, Spotify’s official message-keepers have strung users along. Reading these message boards is a study in corporate indifference. In this board, for example, which began in 2012, one customer after another explains why the ability to alphabetize their list of playlists is necessary if they are to continue using the service. It’s almost comical to watch the customer support folks react to each post as if it is a new idea; or attempt to pacify the customer by assuring her that staff is “working around the clock to implement this feature.” That last comment was made in 2015, three years into the thread; there’s been no word about the feature since.
The desktop player does let users change the order of a given playlist by dragging it up or down. That feature would suffice for someone who had three playlists. It might even work for someone with a dozen playlists. But for someone with several dozen or more playlists, manual drag and drop is not only no solution, it’s actually insulting.
What Spotify has done is create an all-you-can-eat buffet, and equipped its customers with a toothpick in place of a knife and fork or chopsticks.
The problem can’t be that difficult to solve, as Spotify has added alphabetization of playlists to its phone and tablet apps. Yet the desktop, a primary source for folks who listen to music while working, remains as primitive as it was in 2010.
Six years of alternately pretending not to know that your paying customers require a basic tool to manage their subscriptions, and pretending to be working on a solution, shows a basic disregard for the paying customer. Which kind of goes along with a disregard for the working musician, who isn’t exactly getting rich off the Spotify royalties that have replaced CD sales.
Apple Music has rubbed me the wrong way since Apple first crammed it into their increasingly dysfunctional iTunes player (whose poor usability is what drove me to Spotify in the first place). I hate that Apple Music shows up on all my Apple devices, even though I don’t subscribe to it, and even after I’ve turn it off in Settings. In this regard, Apple today is like Microsoft in the 1990s. And I don’t mean that in a good way.
But, as obnoxious and overdesigned as it is, there’s one thing I like about Apple Music: it just may drive the complacent management at Spotify to actually start listening to their customers.
2006 DOESN’T seem forever ago until I remember that we were tracking IE7 bugs, worrying about the RSS feed validator, and viewing Drupal as an accessibility-and-web-standards-positive platform, at the time. Pundits were claiming bad design was good for the web (just as some still do). Joe Clark was critiquing WCAG 2. “An Inconvenient Truth” was playing in theaters, and many folks were surprised to learn that climate change was a thing.
I was writing the second edition of Designing With Web Standards. My daughter, who is about to turn twelve, was about to turn two. My dad suffered a heart attack. (Relax! Ten years later, he is still around and healthy.) A List Apart had just added a job board. “The revolution will be salaried,” we trumpeted.
Preparing for An Event Apart Atlanta, An Event Apart NYC, and An Event Apart Chicago (sponsored by Jewelboxing! RIP) consumed much of my time and energy. Attendees told us these were good shows, and they were, but you would not recognize them as AEA events today—they were much more homespun. “Hey, kids, let’s put on a show!” we used to joke. “My mom will sew the costumes and my dad will build the sets.” (It’s a quotation from a 1940s Andy Hardy movie, not a reflection of our personal views about gender roles.)
Jim Coudal, Jason Fried and I had just launched The Deck, an experiment in unobtrusive, discreet web advertising. Over the next ten years, the ad industry pointedly ignored our experiment, in favor of user tracking, popups, and other anti-patterns. Not entirely coincidentally, my studio had just redesigned the website of Advertising Age, the leading journal of the advertising profession.
Other sites we designed that year included Dictionary.com and Gnu Foods. We also worked on Ma.gnolia, a social bookmarking tool with well-thought-out features like Saved Copies (so you never lost a web page, even if it moved or went offline), Bookmark Ratings, Bookmark Privacy, and Groups. We designed the product for our client and developed many of its features. Rest in peace.
I was reading Adam Greenfield’s Everyware: The Dawning Age of Ubiquitous Computing, a delightfully written text that anticipated and suggested design rules and thinking for our present Internet of Things. It’s a fine book, and one I helped Adam bring to a good publisher. (Clearly, I was itching to break into publishing myself, which I would do with two partners a year or two afterwards.)
In short, it was a year like any other on this wonderful web of ours—full of sound and fury, true, but also rife with innovation and delight.
As part of An Event Apart’s A Decade Apart celebration—commemorating our first ten years as a design and development conference—we asked people we know and love what they were doing professionally ten years ago, in 2006. If you missed parts one, two, three, or four, have a look back.
I’m traveling by train next week. This morning, Amtrak .com wouldn’t let me log into my account with my password, which I have never changed, and which Amtrak.com now says is “incorrect.” When I eventually clicked the “Reset password” link (duh, so Amtrak would send an email with a link allowing me to reset my password), Amtrak thanked me and then never sent that email.
I know they didn’t send that mail. I checked Spam, Trash, Junk, and Junk Mail (I don’t know why my Gmail offers all four of those); ran Gmail searches on amtrak.com; and sent the “Reset password” request several times to be sure Amtrak was definitely not mailing me the “Reset password” email after failing to recognize my password.
Well, when a company no longer accepts your password, and when it will not let you create a new one (because the “Reset password” button does nothing), your next option is simply to give up all your miles and just start a new account, using an alternate email address and creating a new password for that account.
I waited 90 minutes before attempting to do so, in case Amtrak, due to an overloaded gateway or poor engineering, was simply experiencing an extraordinary delay prior to sending “Reset password” emails. Patience is not my chief virtue; anxiety around even familiar, comfortable travel impels me to ward off panic attacks by acting too fast for my own good. But today I did not do that. Today I waited the full 90 minutes.
Then I filled in all the fields at Amtrak.com to create a new account, wrote down my new sign-in information so I’d be able to use it again in the future, and hit “Register.”
At which point Amtrak sent me the following error message:
Due to a system issue, we are currently unable to process your request. Try again later. [Error ID: 1009C]
So my guess is that my original password was probably correct, that Amtrak’s backend simply isn’t functioning correctly, and that when Amtrak can’t connect to its database, it sends incorrect error messages by default, i.e. “your password was wrong” instead of “we cannot connect to our database.”
(I say “backend” and “database” rather than “site” because my Amtrak phone app also bumped me out, and won’t let me sign in. So the problem isn’t specific to the web experience.)
AS MY design career has taken on more and more strategic and managerial freight, I’ve done less and less hands-on design. This year, I decided to change that. As part of my reimmersion, I found myself reading less, and absorbing visual information more. Enter Pinterest.
I’d played with the app when it first came out—who didn’t?—but it didn’t stick with me the way a handful of apps do. It didn’t become an obsession, and so I gradually forgot about it. That’s just how apps work for me. They’re heroin, or they’re nothing.
But the moment my days began filling with sketching, and coding, and Photoshop comping, the genius of Pinterest, and the addictive high it provides when used obsessively and compulsively, was revealed to me.
In borderline religious ecstasy, I became a Pinterest junkie, compelled to collect and catalog every artist I’ve ever loved—every type designer, illustrator, filmmaker, social absurdity, comic book character, and book designer; every half-forgotten cartoonist; every city or nation I’ve visited.
Using Pinterest not only revived long-dead visual design brain cells, it created new ones. Work-related layouts and color schemes came easier as I spent more and more “downtime” collecting and cataloging half-forgotten styles, genres, and artists—and discovering new ones.
I ? Pinterest
As part of this work—for work it is; call it “research” if you prefer—I spent hours rearranging Boards on my profile for maximum aesthetic effect and rhythm. And more hours choosing and replacing the cover illustration for each Board. (If you don’t use Pinterest, here’s a summary: it lets you pin any image you find on the web, or on your own computer desktop or mobile device, to a virtual whiteboard. Pinterest calls each whiteboard you create a “Board,” and each image you affix to it a “Pin.” Part of the fun comes from sequencing Boards on your profile for aesthetic or educational reasons; choosing the featured image for each Board is likewise important and fun.)
Until a few days ago, you could edit and re-edit the featured image for each Board whether you were using Pinterest on the web (that is, via desktop computer), your phone, or your tablet. Doing these things worked differently on the different devices—choosing the featured image was actually faster and less tedious on iPhone and iPad than it was on the web—but the functionality was available in all three places, because Pinterest recognized that brands exist between devices, and that folks interact with your service on different devices at different times, as they choose.
Likewise, until a few days ago, you could change the order of Boards on your profile via drag and drop whether you were using Pinterest on the web or your tablet. (Likely because of screen space constraints, this functionality was not available on iPhone, where the display of Board content necessarily differs from the more desktop-design-focused method used on the web and on iPad.) Users like me changed the order of Boards to create visual interest, set up ironic contrasts, create visual rhythms up and down the screen, and so on. I’m a designer. I have my ways. These details are important to me—and, I imagine, to many other users, since Pinterest is a drug for visual obsessives.
An unexpected change
Then, a few days ago, Pinterest released an update that removed this functionality from the iPhone and iPad (and, I’m assuming, from Android as well). There was no blog post announcing the change. And no rationale offered for taking away features that mattered a lot to users like me. Pinterest knows these features matter, because Pinterest has our data. That’s the difference between making a digital product folks interact with via the internet, and making, say, a toilet plunger. If I manufacture toilet plungers, I can make assumptions about how folks use my product, but I probably don’t have much real data. If I make an application people use via http, I know everything.
Now, it’s not like people were complaining about the ability to edit their Boards: “We have too much freedom! This software provides too many delightful functions. Please remove two of them. But only from my mobile device.”
No. The features are still there on the website. So Pinterest knows people like these features.
And it’s not like the features are too difficult to put into mobile devices, since they already existed in those mobile devices.
A failure to communicate
You may ask why I’m telling you all this instead of telling it to Pinterest. Good question. The answer is, I tried telling Pinterest, but they don’t provide a forum for it. And that is the biggest problem. A company that makes products people love should have a way to communicate with those people. Not grudgingly offer them a few character-limited form fields on a “survey” page that isn’t even referenced in the site’s navigation.
When the features stopped working on my iPhone and iPad, I assumed something had gone wrong with my apps, so I deleted and reinstalled them. (Remember, there was no announcement; but then why would any company announce that it was taking away loved features for no apparent reason?)
When deleting and reinstalling didn’t help, I sought help and contact pages on Pinterest (and was only able to find them via third-party search engine).
In trying to file a bug report, I ended up in a pleasant (but confusing) conversation with a very nice Pinterest employee who explained that I wasn’t experiencing a bug: the software engineers had made a conscious decision to remove the functions I use every day … and had no intention of restoring them. She wasn’t able to tell me why, or point me to a URL that would offer a rationale, but she did tell me I could use Pinterest’s “Recommend a feature” form to “recommend” that the software engineers put those features back.
Since “Recommend a feature” is hidden from site navigation, the kindly person with whom I was in dialog provided a link where I could type in a few characters requesting that Pinterest restore the “drag Board order” functionality. There wasn’t room in the form fields to explain why I thought the feature should be restored, but at least I was able to make the request. The form asked if I was a Business account user, which I am. I don’t remember when or why I bought the Business tier of service. Maybe for the analytics. Maybe just because, as someone who makes stuff myself, I choose to pay for software so I can support the good people who make it, and do what I can to help their product stick around.
(It’s the same reason I remained a Flickr Pro user even after Yahoo gave the whole world 2GB of photo storage space for free. If everything is free, and nobody pays, services you love tend to go away. Half of web history is great services disappearing in the night after investors were dissatisfied with only reasonable profits.)
I don’t know why my paid status mattered to Pinterest, but I couldn’t help feeling there would be a prejudice in favor of my comment if I checked the box letting them know I was a paying customer. Even though it was information they requested, checking the box made me feel dirty. I also wondered why they were asking me. I mean, don’t they know? I gave them the email address they use for my login. I was logged in. They know my status. Are they just checking to see if I know it, too?
There can only be one (feature request)
But I digress. Because here is the main point. The moment I submitted the tiny, inadequate form requesting the restoration of a recently removed feature, the site set a cookie and sent me a message thanking me for completing the “survey.” It wasn’t a survey, but I guess one task completion message is as good as another.
Then I tried to use the inadequate form to report my second concern—the one about the removal of the ability to choose a featured image for my Board. The way this had always worked on the tablet was far superior to the tedious, painstaking way it works on my desktop. On the tablet, you could scroll through all your images with the flick of a finger, select the image you wanted, and complete the task in a few seconds. On the desktop, you had to click your way through every image on your Board in reverse chronological order. It’s the difference between flicking through a calendar, and clicking backwards from today, to yesterday, to the day before yesterday, and so on. The tablet version was fast, easy, intuitive—you interact directly with your content; you can see all relevant content at a glance. The desktop version is cumbersome and 1999-ish. If I had to pick which platform must lose the functionality I relied on, I would not have chosen the tablet. No customer who used the feature in both places would.
But I wasn’t able to share even a few characters of this thought with Pinterest, because once you submit a “survey” requesting a feature, a steel wall in the guise of a cookie slams down, and you cannot make a second feature suggestion.
Not even the next day. (Which is today. Which I just tried.)
This is a love letter
And that is why, as a hardcore fan and user of Pinterest, a service I love and use compulsively, I am using the public web rather than Pinterest’s somewhat unhelpful help center, to share my request with the brilliant software engineers who create this fabulous product.
And with designers, because these are the mistakes we all make when we create products and content sites. We think we are all about the people who use what we create. But we are probably frustrating the pants off them with our arbitrary design decisions and inadequate customer feedback mechanisms.
“You don’t get to decide which platform or device your customers use to access your content: they do.”—
This fetish for pointless customization—customization for its own sake—persisted through the MySpace era, where it actually made sense as an early mass offering of page owner personal branding. Its descendants are the WordPress, Tumblr, and Squarespace themes that create a professional appearance for the websites of individuals and small businesses. This is a positive (and inevitable) evolution, and a perfect denouement for the impulse that began life as “user customization.”
Now notice who else provides some of this same inclusive customization function: the mighty New York Times.
12 LESSONS from An Event Apart San Francisco – ? 1: 
