1. UX Virality Week: 4 of 7

    In real life, like in nature, if your virus can’t adapt to its environment it will not survive as long. You can design that characteristic into your creations so it will change as it spreads. So today we will learn about:

    Designing Adaptability


    (Missed the first three lessons? Start here.)

    In nature, a virus will mutate and adapt as it encounters resistance. The strong versions of it spread more while the weaker versions spread less.

    The same is true of digital viral things. But how can you make them adapt?

    Many people will tell you that being able to personalize your product or features is a good thing, even if those same people usually can’t tell you why.

    Personalization does work when it is done properly, but there are other, more subtle ways to achieve adaptability, so each person has a reason to share your product or content, as if they were the first user.


    Let them customize.

    Real viruses often use your existing biology as a “host” for their greater purpose.

    As a designer, it is better to work with the user’s quirks than against them.

    The most straight-forward version of this idea is to design features that let each person customize the experience to their own needs. 

    The downside of customization is that only power users do it in real life, and it takes extra work, so you have to make it meaningful and valuable. 

    Stardoll (where I worked, once upon a time) is a game community for young girls. The main part of registration is to create a doll; your avatar. i.e. — it represents you. So, of course, everybody makes it look like an ideal version of themselves.

    And you’re not just going to hide that perfect version of yourself in a closet, are you? No! You’re going to show it off and invite your friends to do the same.

    And before you say that example doesn’t apply to you, you should know that the highest-paying customers of all time on Stardoll are adults. And I don’t mean 20-year-olds.


    Create a template for DIY simplicity.

    Real viruses are made of DNA, so they have a built-in ability to replicate and mutate as they spread. You need to design the same quality into your work to make it super-viral.

    The Harlem Shake exploded onto YouTube last year, and everybody wanted to be a part of it.

    Actually it spread because everybody wanted to be a part of it.

    What you may not have realized is that “doing your own” Harlem Shake video only means something if everyone understands what it means to “do” a Harlem Shake video.

    In that case, there was a template or “rules” that were easy to copy:

    One person starts out dancing weird, in a place full of still people dressed up as crazy as possible. Half way through, the music shifts and everybody freaks out. Then, in the final seconds there is a brief moment of slow motion.

    Copy, paste, post, and voilá: A meme was born.


    Open source it. 

    The idea of “open source” is essentially giving away the code or the resources to build something again. It is a way to “give away” a piece of talent, but it is also a way to let people continue a project in a direction that wasn’t originally planned.

    This could be anything from a complex program to a tool that helps you copy memes — just add text!

    Ultimately, the reason open source makes sense in business, is that the original creator often becomes the center of the conversation, if not the provider of the service. You will have hundreds, thousands, or millions of people adding value to the original creation.

    More value, more users. More users, more value.

    That’s a viral structure. 


    Make it easy to parody.

    Imagine yourself doing a “funny version” of your product, or campaign, or trying to list the things that make it recognizable. Is it easy or hard?

    Steve Jobs used to wear the same clothing for every presentation, because it made him a character of sorts. The black turtle neck, jeans and glasses were his look. A kid could go out as Steve Jobs for Halloween pretty easily.

    The Harlem Shake was the same. Buzzfeed quizzes are the same. Gangnam Style is the same. Pharrel’s “Happy” video was the same, and Weird Al DID do a parody of it.


    By designing features or characteristics like these, you allow each user to have a personal reason to share, or to re-create your virus in their own image.


    Tomorrow we will learn why some things with viral characteristics spread like the plague, and some just seem boring: Emotion.


  2. UX Virality Week: 3 of 7

    One part of virality that I rarely see anyone discussing is how the format of the viral thing, and how that changes the way it spreads. The way something spreads also effects the number of people it can reach, so today we will learn about:

    Transmission & K-Factor


    (If you missed the first two lessons, start here.)

    Every tv show or movie that deals with a virus includes a scene where board-room bureaucrats learn how the virus is transmitted from one person to another.

    It could be air-borne, covering a large area but getting weaker with distance. It could be in the blood, which makes it hard to spread, but very effective when it does. Or it could spread through silly dancing at concerts or in nature.

    Viral products and content work the same way, metaphorically.

    Maybe you have to invite a specific person directly — like Dribbble —which spreads more slowly, but probably comes with more credibility.

    Maybe it’s a photo or video that you can post on Facebook or Twitter. Easy to spread and understand, but doesn’t last very long.

    Or maybe using the product itself is viral — like email — where every mail let’s people know that your messages was “sent from Hotmail”.


    Make it very contagious.

    When you get the flu, the best thing you can do is stay home from work, to prevent other people from getting sick.

    When we design a virus, we want the opposite. We want as many people to get it as possible. 

    If every person that gets the flu makes two other people sick, your whole office will become snot factories in a few days.

    If every person only makes 0.5 other people sick (on average) you might not even notice.

    Think about that: the difference between spreading 2 people at-a-time and 0.5 people at-a-time is the difference between an epidemic and nothing.

    That number is called a K-Factor.

    The “tipping point” for a virus is a K-Factor of 1.

    Anything higher than 1 and it will spread exponentially. Anything lower, and it will slowly fade away.

    Generally speaking, the more “invites” or “shares” something gets, the higher the K-factor will be. 


    As a UX designer: your goals will determine which viral strategies are best for you.

    Dribbble is a community where quality matters, and growing super-quickly might actually hurt that quality, so they give more invites to people who post popular stuff. That makes the quality viral, and Dribbble more credible, even though the site grows more slowly.

    Buzzfeed, on the other hand, doesn’t give two shits about quality. They are all about pageviews. So everything they do is based on catchy headlines, funny pictures, and quizzes to see which category of irrelevant nonsense you are.

    Dribbble has grown into a strong, loyal — not-so-big — community of designers that influences trends. Buzzfeed makes craptons of money, but they have no loyalty or credibility at all.


    Making your creation super easy to spread is one thing, but it’s more helpful if you can recruit people to spread your design for you.

    Tomorrow we why letting people change your content or product can make it more viral than any feature you could design: Designing Adaptability. 


  3. UX Virality Week: 2 of 7

    If you want something to be really contagious, ideally it should jump from person to person as fast as possible. In product design, the time it takes to download something and “infect” the next person is called:

    Viral Cycle


    (Did you miss the first lesson? Start here.)

    If you have seen the movie World War Z, you probably remember the scene near the beginning when Brad Pitt first watches someone get bitten and change from normal terrified person, into mindless, pissed-off zombie.

    As señor Pitt counts to 10, the person goes through a full transformation, and he is afraid, because he realizes what that means.

    It has a short viral cycle.

    The scene is specifically meant to illustrate one of the fundamental principles of viral design: the faster it takes effect, the faster it spreads.


    Make the user understand. Quickly.

    When someone first visits your site, or downloads your app, or opens your software, they often to need to set it up, or register, or get a taste of the content, or learn what the buttons do, or play the first level… so they understand the value of it.

    This is why “onboarding” — designing the first experience — is such a big focus for startups.

    This is why an iPhone has a partially-charged battery when you take it out of the box.

    So you get to the good stuff faster.


    Time is money.

    Your first reaction might be to think: “why spend all that effort creating something that doesn’t make the app better and doesn’t make any money? If your product is easy, people will figure it out.”

    Technically, that’s not untrue. If the product itself is good, the user won’t necessarily get more out of it by getting started faster — other than a bit less annoyance.

    But we’re not talking about a user. We’re talking about a virus.

    Imagine two apps. One app that takes 10 minutes to get started, and one that takes 60 minutes. Fairly realistic. Otherwise, assume they are both amazing apps.

    Viruses spread by multiplying exponentially. They infect one person, who infects 2 more, who infects 2 more, and so on.

    That means that the 10-minute app can infect 32 people in the first 60 minutes, while the other app can only infect one person.

    If we’re talking about zombies, a 10-minute viral cycle is scary. The 60-minute viral cycle is one zombie, locked in a room. 


    So when designing something viral, the time it takes to get started and share it with the next person is called the Viral Cycle. And the shorter the better. Small differences for each user can add up to huge differences as it spreads across a network of people.


    Tomorrow we will learn about the way a virus is spread from one person to the next, and how that can effect the contagiousness: Transmission & K-Factor.


  4. UX Virality Week: 1 of 7

    Every day this week we are going to learn something about the ingredients that make one design more viral than another. But what do we mean by “viral”? Well, that’s actually the first lesson:

    The Difference Between Virality and Popularity


    You don’t have to be online very long before you hear or read people talking about something “going viral”.

    What they usually mean is that it became very popular, very quickly. 

    If you’re around startups, everybody thinks their shit is going to be “viral” even when there is no reason to believe that.

    The reason: most people don’t actually understand what “viral” means, and it does not mean “popular”.

    Something can be popular without being viral, like [insert celebrity sex tape scandal here] or [insert latest trendy mobile time-waster game here] or Reddit.

    It might just be good. 

    And something can be viral without getting a billion views on YouTube or being an overnight success, like Pinterest, or email, or Quora.

    Virality often starts slow, but its momentum can become much more powerful.

    (If you’re an investor, I would humbly suggest that you use this week’s lessons as a checklist every time someone pitches you a “viral” business. If it is missing more than a couple of these ingredients, the odds are against them.)

    I have probably already broken your head a little, so let’s look at where the word “viral” comes from, and it will make a little more sense.


    virus is a real thing. I am not talking about a computer virus — which is a program, and often is ”viral” — I am talking about the medical thing. The organisms that invade your body and make you sick.

    A living, multiplying, contagious, adapting, virus.

    There are people that study real viruses, and they know that some viruses are more effective than others, in nature. They can predict how fast a virus will spread, or how dangerous it will be, based on its “design”. 

    Some will spread through a population of people in days. Some spread slowly over a generation.

    Sometimes nobody dies. Sometimes lots of people die.

    Sometimes a virus infects almost everyone. Sometimes only a handful of people are effected.

    So what’s the difference?


    Virality is built on certain characteristics of the virus itself. Or in our case, the product, content, or idea that we create as designers, marketers, or developers.

    Viral things also spread in a somewhat predictable way when you analyze the data — especially on social networks — although it feels unpredictable and crazy while it’s really happening.

    Being viral — regardless of popularity — means your design has viral characteristics; viral structure. Or, it means you’re distributing it through something with viral structure, like Twitter.

    Being popular just means a lot of people like it.


    Over the next six days we will look at several of these viral characteristics, and how you can design them into your work to increase the chance that your creation will spread like Ebola. Hopefully with nicer symptoms. 

    Tomorrow we will learn why your designs should have a short Viral Cycle.

  5. Link: Breaking Web Design Conventions = Breaking the User Experience

    This article is a case study about Bucknell University, where they re-designed their site in an “unconventional” way, causing controversy and heartache for many.

    Boo fucking hoo. 

    I like the Nielsen Norman Group, and generally their blog is one of the best sources of solid UX research, but from time to time they just feel… old.

    This is one of those times. 

    There is a glaring problem with this case study: the data is used to show that the design is bad — which it is — but the blame is placed on the fact that it is unconventional.

    Unconventional design and bad design are not the same thing!

    Controversial design is also not necessarily a bad thing. Sometimes people freak out because the changes are big, not because they are bad. They quickly get used to it, and sometimes like it better than before.

    In a nutshell, the university chose some bad information architecture and it confused users and made it difficult to find the fundamental information that students and prospective students typically want.

    That’s just weak UX work. They prioritized brand over function.

    The fact that the design also happened to be unlike a “typical” university website has nothing to do with it.

    If unconventional design was always a bad thing, every website would look the same, and new features would almost never appear.

    But websites are all different, and different purposes require different designs.

    The designers of the site should have consulted users, rather than prioritizing what the university itself wanted, which — based on their choices — is clearly what happened. 

    Read the article; it’s a good case study, especially for young designers who want to analyze something standard and familiar.

    But your take-away from the case should not be about how conventional the design is. Fuck that. Grow some balls, Nielsen. The data shows that the design was ineffective, not that it was too unconventional.

    Read it now »

  6. Do some science. Find the answer. Look smart for real.
    From the article:
    "UX Diagnostics"

  7. Just because you feel strongly about it, doesn’t mean it matters.
    Base feelings on importance, not vice versa.

  8. The UX of a Sunburn

    As I sit here writing this, I have a really bad sunburn. Looking back at my “last day of vacation” behaviour, I should have expected this. What I didn’t expect was to fix it with UX.

    Day 1 and Day 2 of this sunburn were what you have all probably experienced at one time or another: sensitive, tender, delicate skin that makes you punch people when they touch it. 

    During those days the shoulder strap of my computer bag felt like the soft touch of tree bark and barbed wire.

    As long as I stayed still and wore comfortable clothing, it was tolerable.

    But if you have ever had a bad sunburn, you know that the pain isn’t the worst part.

    The worst part came on Day 3. The itching.

    At 4:30 this morning I woke up in agony, tortured by a thousand continuous, overwhelming prickles as my sunburn caused the layers of my skin to blister.

    You couldn’t see these blisters. They were invisible. But fucking hell, they were in there.

    Basically, I was panicking. I couldn’t stay still. It wouldn’t stop. And it was mostly on my back, where I couldn’t reach.

    I had two possible options. After momentarily considering suicide, I decided to wake up my girlfriend instead, so she could help. 

    While frantically changing clothing, taking cold showers, squirming in bed, and Googling home remedies (white vinegar, who knew?!), I had a moment of clarity about two important aspects of the UX process.


    1) There IS a difference between empathizing with a problem, and understanding a problem.

    My girlfriend, bless her soul, just wanted to help. 

    She assumed I was in pain. After all, I did have a sunburn. It did hurt. was clenching my fists at all times. My whole back was red, for Christ’s sake!

    Not an unreasonable assumption.

    So without a word, she jumped up, got a cold wet towel and put it on my back. She was also careful to touch me gently, so she wouldn’t cause any punching.

    She was super empathetic. She cared. She was emotionally involved. I couldn’t have asked for more effort. Especially at 4:30 in the morning. 

    Problem: I wasn’t in pain. I was really fucking itchy!

    So that cold wet towel started to evaporate, which made the little hairs stand up, and her gentle touching gave me that “oh-my-god-I-am-going-to-break-something” sort of aggression.

    Not good. I think the government should sunburn terrorists and wait three days instead of waterboarding. 

    It wasn’t her fault. After all, on Day 1 and Day 2, that would have felt great. But this was Day 3. My problem had changed.

    She quickly observed that this treatment was making things worse, and started to ask questions. And I started giving feedback. 

    "Don’t be gentle!"

    "Get that fucking towel off me, please!"

    "Yes, you can cover me in vinegar! Just do it!"

    When it clicked for her that the sunburn itself was not the problem, she pulled out some hydrocortizone, which she had for mosquito bites. 

    I have never felt such beautiful, perfect calmness. 

    30 minutes later I was asleep again.


    She was very empathetic to my problem at first. She has had sunburns. She obviously cared that was I was in agony. She wanted to fix it.

    You can only understand a problem after observing it. She tested her assumptions, got feedback from me (the user of the sunburn) and changed her approach to fit the new information.

    She couldn’t see the itching. As designers, we often look for what we can see, or measure, or what people say to us. But that isn’t always the truth we need.


    2) There is a difference between emotions caused by avoiding a problem, and by needing to conquer a problem. 

    This, to me, was a very interesting thing to realize in-the-moment.

    When my problem was pain — something humans want to avoid — my approach, like most people, was to be careful. I was protective. 

    I chose clothing that was soft. I tried not to move that god forsaken computer bag too much. I told people that handshakes were better than bear hugs.

    As long as I did that, the problem didn’t happen, and I was happy. The “experience” of the sunburn wasn’t that bad, actually. 

    However, on Day 3, everything changed. 

    Now I had a constant problem that needed to be conquered, and what was my response?

    Aggresssion. Action. Speed.

    I would have paid $1000 in that moment, just to make it stop. I was desperate.

    So as a UX designer… what type of emotions and behaviour are you seeing?

    Are people angry? Or hesitant? 

    Are they asking for a solution, or are they not doing something in particular?

    Are they desperate to change something, or careful to avoid it?


    Next time you’re designing a solution to a problem, assume nothing. Stop to gather data, ask questions, and observe before you take action. Decide what the problem is, then solve it.

    Don’t make the itch worse by protecting the burn.

  9. Law of Prägnanz: The human brain tends to process simple patterns — regular, even, and orderly—faster than patterns that are more complex.

  10. UX Diagnostics

    There are almost always 10 different ways to create a certain behaviour. That means everything your users do could be caused by something other than what you intuitively think.

    It’s time to play the “what could cause that?” game!

    Maybe your users are quitting after the first visit. Maybe they don’t spend very long on your site. Maybe they don’t trust you.

    There is a good chance you have a theory about the cause of that problem. 

    Maybe there isn’t enough content. Maybe it’s hard to navigate to more than one section. Maybe your design is a little rough.

    There is also a good chance you’re wrong.

    So… what else could it be?

    When you have a theory in mind, one of the hardest things is to assume you’re wrong, or come up with other ideas about what could cause the same problems.


    UX can sometimes feel like the perfect opportunity to be “right” and that feeling is often wrong.

    We all want to be the person with the answers. But it actually shows more intelligence when you realize that there is no way to know the real cause. There are often several realistic explanations for whatever you’re seeing.

    Don’t be afraid to say “I don’t know,” or “it could be several things,” to your boss or your client or yourself!

    Maybe your users don’t come back because they didn’t like what you made them, or because they forget about your site, or because they only need your service once a year.

    Maybe they don’t spend very long on your site because they don’t believe what you say, or because they can’t find the “register” button, or because their friends aren’t there.

    Maybe they don’t trust you because they have never heard of you, or because you don’t sell brands they want, or because you’re adding a bunch of extra costs at the end of the checkout flow.

    Once you have a list of plausible hypotheses, it’s time to start eliminating them, with science!


    UX requires diagnostic thinking.

    Like doctors trying to find out why you’re sick, based on your symptoms. You have a cough, and you’re tired, and your anus hurts? Maybe it’s the flu. Maybe it’s a deflated lung. Maybe you were abducted by aliens.

    The only way to know is to do tests or experiments that will give you more information.

    The flu will have symptoms or treatments or test results that a deflated lung and alien abductions won’t have.

    Do the tests. (A/B tests and user tests)

    Look for the other symptoms. (ask more questions, get more data)

    Try the treatments. (redesign some stuff, see if it helps)

    Whatever happens, you will have more information, and more information in UX means fewer ideas that make sense. i.e. — a better diagnosis.

    Sometimes you should do an A/B test specifically to prove that you’re favourite theory is right. Or to rule out a possible cause. Or to show a client that their theory is unlikely to be the real reason.

    Or maybe they are right! Shit happens! 

    Instead of picking one hypothesis and trying to make everyone believe it, try testing instead. When you prove the right answer, you will build respect among your colleagues, you will learn and improve yourself, and most importantly, you will know that aliens are out there, and they do use anal probes.

    You will also understand your users better than you did before. if you’re into that sort of thing.


    Do some science. Find the answer. Look smart for real.

    You might surprise yourself!