Powzot

Sneaky things you can do to make your designs "Human Centered"

25 Apr 2011
0 Responses
views
Interaction Design Research UX Usability

Saturday I read a fantastic article by Whitney Hess, "You're not a user experience designer if..." (and you should read it too). In it she brings up really excellent points about things that a lot of "UX designers" don't actually do. Things like: talk to users, design in a vacuum, make design decisions based on personal preferences, don't use UX methods, etc. While I completely agree, and shouted it from the Twitter roof-tops, it was a part at the end that caught my attention and made me decide to write this blog post.

If you have the title of User Experience Designer and you want to do these things but aren’t being allowed to, don’t stand for it.

I've been in that situation. A lot. In fact, it was not being able to do these things as an "interactive designer" or "interactive art director"(titles that I held for the first 4-5 years of my career) that caused me to almost give up on design completely, consequently find this whole thing called "UX Design" and figure out that I was trying to do "UX design" in addition to the graphic and interaction design I was doing. It's because of this that I passionately argue for real UX design. It's also because of this that I've hacked together some covert UX methods that you can use if you're in a situation where you want to do real UX, but can't.

First, a disclaimer

These are, more or less, methods I've used over the last few years when I just couldn't get approval to do real user research or testing. I will never, ever ever ever say these are "good" methods, they're just better than not doing anything. These are in no way "science" and are so far off from pure UX design methods that I'll probably get some nasty emails and comments anyway. Still, as I've been told by multiple people I really respect, "Any research/testing is better than none."

Also, I take no responsibility if you get fired because you get caught using these methods. I'm sorry, but it was probably for the best anyway.

Sneaky things you can do to make your designs "Human Centered"

Informal happy-hour individual interviews

Chances are you probably know some people on Twitter/Facebook/LinkedIn/In Real Life that are like your users. (at least in someway) Can't do a real ethnographic study? Can't do formal, structured/semi-structured interviews even over the phone? It's time to head to the bar! Pick a few people and a few days, and individually take them out for a drink, preferably someplace where the music isn't too loud. Bring a notebook (or use an audio recorder if you can get away with it) and ask them questions. Ask them their opinions. TALK TO THEM. Try not to drink too much (at least until you're done.) The beauty of this is two fold. First, your participant's incentive is built in. Secondly, you can often times get a 2nd informal interview with someone that is in the bar who overhears you. Is there a sampling bias? OH YEAH, but if they pass your screening criteria take what you can get. More data is better than less in these situations.

Surveymonkey after-hours survey

Surveys are great, because they can kinda run themselves now thanks to the internet, all you have to do is write them. They're also great because you can get a 10 question survey with up to 100 responses per month for FREE, which means this one isn't actually going to cost you anything. (No this isn't a paid endorsement) 100 responses is good for 90% confidence (p=0.1) just FYI, so your results will actually have some statistical validity too! (assuming your questions are good)

Diary study with friends or others who will help you

So you can't go and run a full fledged ethnographic study, if you have really awesome friends who are close to your target user group you can get them to take a cheap notebook to work with them and self-report the things they do/frustrations they have/whatever you need to find out. Is it as good as observing? Nope. You're gonna miss a lot, I'm not going to lie. However, you're not going to miss as much as you would not doing it and just guessing.

Informal happy-hour cognitive walkthrough

This is a lot like the happy hour individual interviews, except you pull some of your coworkers instead. Bring your wires, or comps or whatever you need to test with you, and tell them you want their opinion on them. The key to this one is informal, don't ask for "stories about using X." This also has the benefit of slowing indoctrinating coworkers into the HCD fold, and slowly changing the culture to be more friendly to all of this stuff.

Paper prototyping on coworkers/friends/family/innocent bystanders

If they're close to your user base, draw them a quick marker comp and ask them to pretend it's your product and perform a task on it. (I've done this on napkins before, which is why it's always good to have at least one sharpie with you at all times.) This is a legit method (maybe the most legit in the list) when you want to quickly confirm a wacky idea without having to spend time doing nice looking wires or comps. I recommend this testing method for anyone working on an early stage of a project.

After-hours card sort

The Internet has made so many things better (like autotuning every sound in nature) including card sorts. Just like the survey, you can run an online card sort and kinda just ignore it except when you're setting it up and analyzing the results. Websort.Net is by far my favorite card sorting tool, and it's totally free.

The bottom line

The real secret is that you can do almost any formal research or testing method informally, under the radar and on the cheap using your social network, assuming you're willing to put in some extra work after-hours. Is it perfect? No. Is it science? Kinda...it's probably got too many confounds to count though. But, it's better than nothing. Unless it's really really small, your network of friends and family online and off probably has at least a semi-representative sample of your user-base.

The real question is how to you present your results, or use it to argue your points? If you can get extra points for, "I did user research/testing in my off time" use it like you would if you had been allowed to do it. If not, just credit your awesome rockstar design skills. Say something like "You know, I just really get people" ...and then start looking for a real UX job.

Ohhhhh you said the F worrrrddd!!!

15 Apr 2011
0 Responses
views
Design Rant Research

It's very funny (to me anyway) that I came out of college decrying research, it "ruined design," and a few years later my love for research informed design pushed me back into college to learn more about it. It's also funny (probably still just to me) that I continue to run into these two seemingly opposed design camps all the time, the pro and anti research designers. I'm among friends in one, and inevitably find myself defending the dreaded F**cus Gr**p in the other. It's like a bad word in a lot of circles, f**cus gr**p, and as I always end up explaining, "it's because people don't run the dreaded f**cus gr**p right." They don't ask the participants the right questions. They use it to answer the wrong kind of research questions. They blindly follow the majority opinion as if it was handed down on stone tablets. Bottom line, they fuck it up. Worse yet, this is the only research method a lot of designers, including fresh out of design school me, know. At the very least the only one they think of when someone says "research."

Research isn't a bad word for designers (even f**cus gr**p)

This is the basic argument I find myself making when in the company of the anti-research designers. I'm a designer, not a researcher. I have a heavy-weight, framed piece of paper that says just that, and in December I'll have two. (This is the only "trust me, I'm a doctor" style statement you're going to get from me, I promise.) As a designer, research is just another tool in my toolbox for solving design problems. It's just like photoshop, or illustrator or my sketchbook. Research does NOT dictate design decisions, it informs them. This is something I usually have to repeat, so I will in an obnoxious type treatment to make it stand out

RESEARCH DOES NOT DICTATE DESIGN DECISIONS, IT INFORMS THEM.

(Sorry, I it wouldn't let me make it blink too.) If you're a designer and you think back to your design classes, you'll likely remember doing some kind of asking around or looking at what other people had already done in that space for every project you did. That's research. AND it isn't a f**cus gr**p. (It's likely a competitive analysis or an unstructured individual interview, for the researchers in the audience keeping track.) Your research didn't totally take away all your choices as a designer, and you probably didn't completely change your initial gut feelings based on it. It didn't dictate your design. But it did inform you, and you might have modified some of your initial assumptions, and you may have abandoned bad ideas if they proved to be bad, or already done, or whatever you found. It may have even given you an idea because you saw something new and it sparked something awesome in your crazy designer brain. This is why good research, used correctly is magical.

Let's talk about the damn f**cus gr**p for a minute

(Warning: this is a rant, you have been warned.)

Most people (especially in advertising) ruin focus groups. This is not an "ivory tower" opinion. This is my on the ground, in the trenches throwing grenades, practitioner's opinion. They're not looking for additional information to inform the design, they're looking for truth. Worse yet, they're not just looking for a truth, they're looking for "The Truth," a totally generalizable (ecologically valid for you researches still keeping track) truth. One that describes everyone like their group everywhere. A group that is usually structured to represent everyone on the planet. Sorry, a f**cus gr**p is the wrong test for finding "The Truth." It's a qualitative test that is at best going to give you an idea of your specific group's opinions, if and only if you recruit the right group and ignore the idea of group think and follow the leader mentality in groups like this, and dont' have one guy who does 80% of the talking for the whole group. At best you're going to be able to say "we showed it to 10 people and 2 of them told us all these things." F**cus gr**ps are not generalizable, period! The are not "The Truth," and to use them as such is foolish, and you should feel sorry for those people because they're likely going to lose a lot of money.

If you're going to hate the f**cus gr**p , hate it for these reasons. Don't hate it because "research kills creativity." it doesn't, it focuses creativity, and gives you new jumping off points if it's done well.

Design is not art, deal with it.

This is the other argument that is used to try and defend the anti-research designer position, "Design is art." This is a horse that's been beaten to death, and I'm usually one of the people holding a bat. Design is not art. It can be objectively evaluated because the whole goal of design is problem solving. The goal of Art is expressing the ideas, emotion and opinions of the artist. You can't do that wrong. That's what makes it art. You can't look at art and realistically say, "The artist totally failed to capture the way she was feeling in this piece." You can look at design reasonably say, "this piece fails to capture the attributes of our brand," or "this product is hard to use and it's even harder to look at," or even "why would you design a shirt with no neck hole, no one can wear this!" Design can be wrong. The confusing part is that there are usually many solutions to the same design problem, and all of them are right, but the wrong ones are still wrong.

There is absolutely an art to design, and great design often does the things art does too, but design is not art. It's design. Even when it conveys emotion. Even when it conveys opinion. Design is concerned with solving a problem that most likely has multiple solutions, and this is why research is such a great tool. It can help you sift through those worse solutions, that are still technically right, to find the great ones. Why wouldn't you want something to help you sift through an endless pile of hay to find the five needles you're looking for?

So what...?

Research is your friend, dear fellow designers. It can lead you to new ideas, better ideas, by pointing to the right path in the tangled map of options we have; and if it's done well that's all it will do. Research can't ever substitute the thoughtful work of a good designer, because good research isn't creative. To quote Dr. David Evans of Psycster research (one of my graduate school instructors):

A research study by it’s very nature is a cheap, fast proxy to reality.

It's a peak out the window. A designer takes all the information she has about a problem, combines it with everything else she knows and turns it into a solution. Well done research is just one more thing the designer gets to know and use. If you want to be an anti-research designer, why not be an anti-bad-research designer?

Are you training your users for a bad experience?

6 Apr 2011
1 Response
views
Interaction Design UX

If you have an email address, you've probably received at least one email this week about how it was leaked by Epsilon (I've received four now.) If you're not familiar with who they are or why they have your email address to lose (I wasn't) Cnet has a pretty good article about it. What I find so troubling about this in particular, is that Chase bank was one of the notification emails I got. This is troubling because of the way Chase's rewards program works, and the behavior they've trained their users to exhibit is now the very same behavior they are warning them against. This is part of experience design that often gets ignored, and now it's going to be huge.

Training your users to do the wrong thing

If you're not a chase account holder, you're likely not familiar with their rewards program and the behavior I'm talking about. Once a quarter, the bank chooses three to four categories that cardholders get 5% cash-back on. They announce this with emails, and if all the emails were doing was announcing these special categories I wouldn't be writing this post. To qualify for this offer, you have to follow a link in the email and sign in at a microsite. Hello fishers, I hope you have your pirated copies of photoshop warmed up.

This is the behavior the company endorsed, following a link in an email and signing in. I'm sure no one in the marketing department, or at Epsilon ever dreamed they would now be sending out emails telling people to do the opposite (don't give your account info to anyone who asks for it via email), but the damage is already done. The mental model is already set for thousands (or maybe even millions), "To get my 5% cash back on X this quarter, I click the link in my email and log in to sign up."

Fish, meet barrel.

How can we do better?

When we're designing systems, any systems, we need to take into account the behaviors we're training users to exhibit. More importantly, we need to be aware of the big red flag, don't ever do that style behaviors for a particular domain. Training users to follow email links to sign in is a dangerous design because it's so easy to spoof an email and a website these days. To add insult to injury, it's completely unnecessary to make people do this. You can track the success of your email by setting a special link to the account site, where there is significant anti-spoofing measures is place, instead. You could just send an email announcing it, telling people to go log into their account to activate it. Or you could just automate it. Asking users to follow a link from an email and sign-in is one step removed from sending them an email and asking them to reply with their credentials, and it's not a big enough step to make it ok.

I'm sure there was a business goal driving this dangerous design, but how much money are you saving now if even 1% of your effected users get fished? How does that experience effect your brand? How does it effect your current customer base of non-fished users? Word of mouth is a powerful marketing force, and it's directly tied to the experience your customers have with your stuff (branding, advertising, customer service reps, third party vendors, and even people you've opened the door for who impersonate you well.)

So what...?

As designers, sure we have to be champions of good design, but we also have a responsibility to avoid designing things that will get us shanked later on. In this case, an already serious data breech is made even worse by a bad design. Chase trained their users to walk right into a fishing scheme, for whatever business reasons. I've yet to meet an interaction or experience designer who isn't really Internet savvy, so I'm sure whoever came up with the model Chase went with was aware of all the things you're never supposed to do from an email. But there is always the chance a designer had no hand in this system, or that they were ignored. We'll never know unless Chase decides to tell us. What we need to do, as a community of design professionals, is keep this event in mind and make sure we're not reinforcing bad web behavior with our systems.

Sneaky things you can do to make your designs "Human Centered"

Ohhhhh you said the F worrrrddd!!!

Are you training your users for a bad experience?

About

Archive

Get Updates

My Sites