This is the transcript for The Dirt episode When Everyone Is Your Target Audience
Steve Hickey: He’s going to breeze right by that before I have a chance
to countermand what you just said.
Tim Wright: Hello, and welcome to The Dirt. I am Tim Wright and today I’m
here with Steve Hickey.
Steve Hickey: Hello.
Tim Wright: And Mark Grambau all the way from Colorado.
Mark Grambau: Hey Tim.
Tim Wright: Hey Mark. How’s your pseudo vacation?
Mark Grambau: My pseudo vacation is pseudo excellent.
Tim Wright: Excellent. That’s good to hear.
Mark Grambau: Thanks.
Tim Wright: This week’s episode of The Dirt is brought to you by [sponsor]
an online scheduling software platform. You can find more
at FTS.io/[sponsor]. We’re super excited to have one of our
first sponsors, our second sponsor.
Steve Hickey: I feel like a real podcast.
Tim Wright: We are. We are totally validated in our whining.
Mark Grambau: That’s right. We’ll go off from just a few guys whining at
the microphones to a few guys whining at the microphones
with a sponsor.
Tim Wright: Yes, I like it and we enjoy [sponsor]. We use [sponsor] at
Fresh Tilled Soil, it’s really cool, and you should check
it out. Today, we are going to talk about something that is
sort of an issue on every project in defining your target
audience, but usually you have clients or front projects
where you’ll say, oh, my target audience is everyone, which
is usually a lie.
Steve Hickey: Yeah, so what we’re actually talking about today is being
specifically non-specific about your audience. The worst
thing you can be.
Tim Wright: Yeah, we’ve worked on some, I guess you can call them
public sector projects where the target audience is
legitimately everybody and it’s a really weird concept and
it affects a lot of the points of the project where you can
target, if you want to target Justin Bieber fans, but we
can’t just target Justin Bieber fans with this because
there’s Justin Bieber fans and there’s Ann Tara fans and
we’ve got to take everybody into account.
Steve Hickey: Yeah, if you’re going to target Justin Bieber fans, you
know a few things like you can only use no more than five
letter words and very simple primary colors, and things
Tim Wright: Bling tags.
Steve Hickey: They’d probably respond well to bling tags.
Tim Wright: Marquee, the old marquee tag. Anyway, there are some UX/UI
design implications and development implications of
targeting everybody and we wanted to talk about that and
how on a recent project we dealt with it personally. I know
you guys being Mark and Steve were specifically on the
project and you guys ran usability tests and did multiple
iterations on it based on user feedback. Can you talk about
some of it on a pod cast?
Mark Grambau: Well, let’s think. It was a project that as we’ve
discussed previously vaguely. It’s a mapping project for a
fairly major sort of public resource, let’s say and the
assets in using a lot of different ways. You’ve got to be
able to use the maps online, using them in actual size that
you see there physically which means you’ve got a lot of
different use cases, so not only is it this major public
location but on top of it you’ve got people looking at this
artwork and looking at this interface from every which
angle of device you can imagine. A lot of our testing went
into legibility from certain distances, ensuring that we’re
not over dependent on texts but also ultimately not over
dependent on iconography.
Steve Hickey: Specifically in that area, there’s a lot of duplication
that we dealt with. We’re talking about an environment
where you can encounter people of just about any age but
also many, many, many different languages so if somebody
doesn’t speak English then the symbol has to make sense to
them in some context.
Mark Grambau: Yeah, and this was a project where international was
targeted but not in a normal sense where I’m from France
and I’m accessing this page from France. It’s I’m in this
Steve Hickey: This already confusing location.
Mark Grambau: Yeah, and I’m standing next to somebody who maybe speaks
English really well and we both have to be able to
understand this product.
Steve Hickey: In an online sense maybe we can eventually localize this
product but in physical locations, you really just can’t
say, oh, here’s a little thing above the monitor and it’s
getting that person. Oh, they speak French. That would be
nice but clearly impossible.
Mark Grambau: That would be profiling.
Steve Hickey: Yeah, I think it would be profiling. We were able to rely
on, there’s an international icon set that is fairly
standardized. I think that everybody who deals with it has
this impulse to go and modify pieces of it to make it
better but I don’t know if you’re really making it better
when it is in fact considered to be universal at this
point, so by relying largely on those icons, I think we’re
able to solve several of our problems.
Tim Wright: Is that an actual icon set?
Steve Hickey: Yeah, it is an international standard icon set. The AIGA
hosts it on their website if you’re looking for it and
these symbols have been tested and used for decades,
particularly in public transportation so they tend to hold
up pretty well.
Tim Wright: Right, so a project like that isn’t the project for trying
out new interfaces?
Steve Hickey: I would say absolutely not. There are things that you can
do that are new and interesting but largely you need to
stick to tested methods and ideas, the icons being one. You
have to have every clear affordances in anything you’re
working with for the interactive components. You don’t get
to be like iOS, screwing around with stuff that isn’t
easily indicated as clickable.
Mark Grambau: I would add that nine times out of 10, yeah, you want to
stick to predictable patterns. However, I would never tie a
designer’s hand. We try not to tie our own to say you can’t
try and invent some great new way of interacting but the
important thing is that we do across all kinds of sign is
building redundancy. If you are going to build some great
new way of doing this as gesture or who knows what that you
also have a clear, visual understanding of how that’s
working and there are other ways of accomplishing that same
I think my favorite classic example of strong redundancy in user
interface design is just copy and paste on your computer.
If you are really into your computer and someone is
teaching you how to do this, you don’t right click, you
don’t have keyboard commands. You just want to select text,
and the smart thing to do for this kind of user is select
the text, go into edit, and you see copy, paste. You have
all those menu items there. For the slightly more advanced
user, they know, hey, I can select this and I can right
click and I can see a contextual menu and I can copy and
paste. Then for the advanced user, they know, oh look I can
hit command c on a windows control C, control V, and now
that kind of redundancy is the way you can explore a little
bit more on your UI.
Tim Wright: Yeah, I think and the project that specific project, we
couldn’t try out something like, Steve you mentioned
something yesterday about progressive reduction or somebody
I forget who brought that up.
Steve Hickey: Yeah, so LayerVault uses a strategy called progressive
reduction where they can because you’re logged into your
account they can collect statistics over time about the
parts of the interface that you’ve directly encountered,
how much you use them, and make a family educated guess on
how experienced you are, and as you gain more experience
with the interface they can declutter it to a certain
extent. Maybe if you don’t log in for a couple months they
might add a few pieces back in the interface, so instead of
something that was text label and an icon before it becomes
just an icon so they can smooth things out.
I don’t think that was really an option we faced here, this
particular product you might only use it once in a while,
in a very great while. Certainly, we don’t have any log in
states so these no way for us to easily remember who you
are without resorting to cookies or local storage or other
unreliable features so this has to always be as clear and
concise as possible.
Tim Wright: It’s just not like something someone is signing into so a
lot of times it’s going to be on a kiosk.
Steve Hickey: Yeah, it’s essentially kiosk-like, think of it that way.
Tim Wright: They’re designing redundancies into the interface making
four different ways to do the same thing but when the
product ultimately goes online, if there was an
authentication system we can do something where we start
pulling away those more, I don’t want to call them dumb
features but redundant, extra, redundant features.
Steve Hickey: They’re not even necessarily redundant. I think designers
love icons because they give them an opportunity to draw
something small and pretty and interesting consistent but
the thing is icons don’t read as well as text labels.
Tim Wright: No, they don’t.
Steve Hickey: People think text labels are boring but they are
definitively better and in this case we had to use text
labels and icons because while the text labels are clear if
you speak the language, there needs to be a fall back for
people who don’t speak the language, so there really is no
opportunity to reduce at that point. We just have to make a
very clear association between the icon text label while
also maintaining a clear hierarchy between different items
Tim Wright: Yeah, and I think we could do an entire show on
progressive reduction and experience decay, but maybe
that’s this show. I don’t know, guys, we don’t know where
we’re going. We never do. But there’s targeting everybody
was definitely a problem but I think it actually helped in
the usability testing stage because we know that we’re
targeting everyone. And one of the things that I’ve said in
the past what I don’t like about the classic usability test
is that you pull someone off the street who has no context
and really it probably isn’t in your target audience to run
your test, but then yeah you find some interface issues
with it, but it’s certainly going to be more useful if you
can get someone that’s in your target audience to do the
test and having such a wide net for our target on this
project landed really well to picking out random people.
Steve Hickey: Yeah, in this case the likely audience is a confused
person who has been thrust into a situation that they have
no prior experience with.
Tim Wright: Yeah, like, waking up on somebody’s lawn without any
Steve Hickey: Yeah, I would like to think that a well-developed and well-
designed sign would help somebody in that situation. A
little home sign and an arrow with number of miles pops up
next to you. You could probably pay somebody to do that.
Tim Wright: We actually had people come into the office and do, you
guys were in the test more than I was. We set up like a
little mock lab.
Steve Hickey: Yeah, so we had a set of printouts and we physically
divided the room up so you wouldn’t see the other printouts
before it was time to expose you to them and we developed a
likely scenario for this person to be going through and
guided them through the different steps where bringing them
to the physical items they would use to solve these
scenarios and see how they react. I think it went very
well. We learned a great deal about what we had done wrong
essentially from the beginning and it was very helpful.
Tim Wright: There was a lot, looking out from the inside, it looked
like there was a lot of information architecture changes
after the usability test.
Mark Grambau: Hierarchy I think was an issue.
Steve Hickey: There’s these items in question had both a key and a
visual display and both of them were having hierarchy
issues where you couldn’t distinguish different types of
items in the key and when you looked at the display which
the key was matched to, visually, many different layers of
information had the same visual weight and that helped us
realize that was a problem when we went back and we created
very clear separations, the things that were absolutely of
vital information to clearing this space and here are the
things of secondary importance and here are the things that
you might never need.
Tim Wright: Yeah, I think from not in this specific project but a
designing for everybody attitude in building this stuff, I
would pay more attention to making things work without
would probably even take it one step further in making sure
that all Ajax calls fall back to actual content and you
know you have to be super accessible in all points. It’s
not that big a deal on a kiosk because no one’s walking up
to the kiosk hitting, what do you have to hit? A control
five times and then you get a windows prompt and then you
have to–yeah it’s a thing. If they give you a keyboard, a
lot of times they’ll block out the control key. If you just
sit there and start hitting the…
Steve Hickey: Is that, I remember hitting sticky keys a lot by accident
and always being infuriated, but I didn’t realize you could
a windows prompt and then you can get to the browser,
Steve Hickey: Oh, oh, I get it so you can get into the back end or the
forbidden area of a kiosk by pounding the control key?
Tim Wright: Yeah.
Steve Hickey: That’s good to know.
Tim Wright: If it’s a window’s machine. I don’t know how, I don’t
remember how I learned that and I don’t even know if you
can still do it but you used to be able to. On a kiosk, no
there is certainly no screen reader in there for folks,
people with disabilities, but in a normal environment, non-
kiosk environment, you’re going to pay extra, extra special
attention to that kind of think.
Steve Hickey: There are many different scenarios under which you can
encounter the product we worked on where there are
accessibility concerns, taken and people will be able to
use it without an issue. However, one of the things that
you are going to find is if you use it on your personal
requirement for this project, which is not something we are
tremendously happy about but a great number of the
technologies we were working with were just not usable
Tim Wright: But there’s fallback content for that because I was–
Steve Hickey: Yeah, there’s fallback content for that. It’s just not
Tim Wright: Because of the nature of the project, you can fallback to
which is not awesome but certainly workable.
Tim Wright: Well, with the progressive enhancement goal of always
keeping the content available, if my goal is to go here and
consume the content, it doesn’t technically matter if it’s
rendered SVG, zoomable SVG or a PDF as long as I’m getting
Steve Hickey: It’s certainly still present.
Tim Wright: Not to, degrade the importance of a good interface but if
the goal is give me my content and you give the content,
then the goal is met and from the get go you’re giving an
acceptable experience across the board.
Something else I wanted to bring up about the LayerVault’s
progressive reduction. The example they gave which we
eluded to was a button with an icon and text and then as
you start using the button the text goes away and then as
you start using the button without the text, it degrades
even further and I saw absolutely no point in doing that.
Steve Hickey: Yeah, I liked the philosophy but honestly with a product
like that I would be way happier if I just always had a
text label. I understand the impulse to just do what they
did but I think it sort of comes from a caring more about
how it looks than how it works perspective and honestly I
haven’t used LayerVault. I know a lot of people like it so
I can’t really know it without understanding how it works
but the way they described it kind of sounds like a selfish
inherence to I want to make icons.
Tim Wright: There’s the two way complexity in, there’s the progressive
reduction going down, removing the treatment to the button
and then there’s the experience decay where if I don’t sign
in for a few months the button rebuilds itself.
Steve Hickey: I think that’s how they said it works. I would assume
that’s how it would work because that would indicate I
haven’t been here in a while. I probably don’t remember
Tim Wright: Yeah, and I think you do have to do that if you break it down,
you need to build it back up. It’s not just progressive
production, let’s clean up the interface. It’s let’s clean
up the interface but also we need to rebuild the interface
programmatically. There’s a lot of complexity built into
something like that.
Mark Grambau: It’s a pretty neat way to avoid the issue of the
inevitable thing that if you don’t keep up with something
constantly you’re going to forget. Folks who learn
Photoshop and then don’t really touch it for six months or
a year, it’s a really nice solution to make it usable for
them again where any of those simplifications back off a
little bit and gives you not quite a first time user
experience but just enough to sort of jog your memory.
Tim Wright: I think it’s actually a good way to stop arguments about
should we have the icon text or not or should we style the
button like this. I feel like a program was just sitting
down like, how about for these users we do this and then
for these users we do this and we just nailed the system.
It’s like a super elaborate system too.
Steve Hickey: Yeah, the thing is we can’t actually know their rationale
and whether or not this is a good idea without having
access to presumably the question they’ve asked which is
does reducing the interface make it easier to use for the
power users at all. If it doesn’t actually improve the
experience for people who are using this thing all day
every day then there truly is no point to it but if it is
improving the experience, if it is allowing people to do
things faster then, okay, awesome.
Tim Wright: I would actually argue that it’s initially not because the
button changes and you have to be like, oh, process it for
a second, the button changed. Okay, now I got a, it’s even
just that little stutter of a few seconds.
Steve Hickey: I’m more interested in how they learn when a person has
done well enough that they can start removing stuff.
Tim Wright: Yeah, well maybe we can get them on a show and talk about
it. I think we could do a how show on progressive reduction
and I don’t want to . . .
Mark Grambau: They have a follow-up blog post after that first one that
was talking about implementing. It’s mostly development
post but it’s showing development wise how they implement
it. That may offer some clues as to the timing on when they
determine someone is experienced enough to get to the next
Tim Wright: We have two events coming up at Fresh Tilled Soil. We are
hosting Global Accessibility Awareness Day here in the
Watertown office on May 15, details to come and Steve and I
will be in Vegas spouting our beliefs.
Steve Hickey: Beliefs, lies, slander, the usual. That will be in June.
Tim Wright: Yeah so we hope to see you in June in Vegas at Future
Insights. Come out and see us and give us hugs.
Steve Hickey: Give Tim hugs. I don’t want sweaty Vegas hugs.
Tim Wright: I do. And of course we want to thank our sponsor this
week, [sponsor] for helping us out with the show and you
can find more about [sponsor] at FTS.io/[sponsor]. As
usual, you can get us on Twitter @thedirtshow and please
send your questions and comments to us there or at
firstname.lastname@example.org, I’m still learning that email
address. Please review us in iTunes or your podcasting app
of choice. That’s all we have for today. Thank you for
listening and we will try and do better next time.
That’s not a problem though. When people using weird, offensive class
names or variables is not a problem.
Steve Hickey: This guy seemed to think it was.
Tim Wright: It’s not. It isn’t. It’s only a problem from the
standpoint of we want to use meaningful variable names.