Re: [agile-usability] The simplest thing that could possibly be re-invented?
- On Fri, 2004-10-08 at 17:45, jeff.grover@... wrote:
>I think that principle is one people often misunderstand, because they
> Intellectually, there is a great temptation to say the "simplest thing
> that could possibly work" (XP/agile principle) is to do what someone
> else held in high regard (OS vendor, competing product, open-source
> project, well-used application) chose to do.
confuse "simple" with "easy". The easy thing to do is often to say,
"make it just like X". But the reason they just point to X rather than
saying what specifically they want is that X is not simple.
Figuring out what is really the simplest thing that could possibly work
often takes a little more thinking, as you have to understand both the
need and the medium in enough detail so that you can pick out the simple
solution that addresses the core need.
> And, more importantly, if I'm trying to be "agile"... how can I argueI think of most moves toward agility as tightening feedback loops. The
> the need for more rigorous user/interaction design and testing
> (feedback), when someone else has presumably already paid for that
> (...or got really lucky and got it right the first time)?
best way to get user feedback is to release early and often while
listening to your users. When you can't do that, user testing can be an
adequate substitute. Either way, getting frequent user feedback is very
much in tune with the agile spirit.
Personally, I think short iterations and frequent releases make it
easier to innovate, as you get
* more data -- prototypes are good, but there's nothing better
than trying the real interface on real people.
* more design time -- agile methods allow you to overlap the
design and construction work.
* less pressure -- when design is no longer a phase, it stops
being a bottlneck. That means less pressure, making it easier
to be creative.
* less risk -- by taking things one step at a time, your bets are
smaller, which reduces the incentive to chose the safe,
* better feedback -- because you can find out how you're doing at
each step, you have a better environment for learning how to
make good choices.
> How can I make it more obvious to feature-driven stakeholders thatI think you need to make this argument in financial terms. In essence,
> there is inherent value in re-thinking and innovating even though
> pre-canned answers to tough usability questions seem there for the
> taking (however false they may be)?
you're saying that you want them to spend their money differently, so
you have to show them how that will result in reduced costs, better
sales, or something else that improves the bottom line.
- Jeff Grover wrote:
... a great deal of the usability input we get whether
"evolutionary" or "revolutionary" originates from or is
relative to products competing with ours or performing
similar functions in some other market space. Here are
"We need a <Fill in your favorite object here> editor
that looks like <our competition>"
"Make the browsing feel like the Windows Explorer".
"Make the web page look like My Yahoo!, Amazon,...
<fill in your favorite site here>"
How can I more eloquently describe this to people
driven by "industry standards", "competetive analysis",
"code reuse", and "feature envy"? And, more
importantly, if I'm trying to be "agile"... how can I
argue the need for more rigorous user/interaction
design and testing (feedback)
I find that when designs and solutions are expressed
relative to another product, it is often the sign of
sloppy thinking. If someone has thought through a
problem, there will be less need for comparison--and a
good chance that he or she will have some drawings,
documents, specifications--anything--that will express
that thinking more completely.
Your examples above are solution statements--divorced
from any particular problem statement. Try seeking the
problem. (We had a good long thread on this a while
ago.) Ask a simple question: "Why?" This may reveal a
lack of thought, or it may yield a very good answer.
Sometimes, when you encounter a lack of thought, you
will find at it's core a lack of understanding of the
problem space. This is where you may have a good
opportunity to employ a more rigorous cycle of user
research, interaction design and testing.
The potential upside is not "innovation" for its own
sake, but rather a solution that better fits the
*specific* problem you are trying to solve.
- --- In firstname.lastname@example.org, "Joshua Seiden"
>I can only agree with what's been posted so far. If there isn't a
>> Jeff Grover wrote:
>> ... a great deal of the usability input we get whether
>> "evolutionary" or "revolutionary" originates from or is
>> relative to products competing with ours or performing
>> similar functions in some other market space.
> The potential upside is not "innovation" for its own
> sake, but rather a solution that better fits the
> *specific* problem you are trying to solve.
financial reason for changing the software, it's going to be a hard
I'm hearing that you believe a usability person might come up with
better solutions than the "us-too" choices you see being made. No
offense to the usability people and designers on the list - but what
they're doing isn't so innovative - not usually. It's exactly what
Josh has said in his post: it's correctly identifying the goals of
the user and determining the simplest way to address those goals.
It's just remarkable how often people don't choose to identify the
goals of the user.
Ok - it's really not quite that simple. You sorta need to know
something about that user and the context of use as well. A little
experience with a few proven approaches doesn't hurt either. But
knowing just a little about user, goals, and context of use can go a
long way at disqualifying "us-too" solutions. The idea isn't to
disqualify them because they're not innovative - but to disqualify
them because they're inapropriate... inapropriate for the goals of
the user, their experience level and the context of use.
Start with identifying goals first. Ask what the user's goals are
until you get a good answer. Try the "popping the why stack"
or "poking with the why-stick" technique. If you really know the
goal of the feature, you often find the simplest appropriate solution
is also the cheapest to build.
Finally, consider trying to make a case for low cost usability
testing. By that I mean recording users actually trying to use the
software - using software that records the screen and the face of the
user running it. There are inexpensive tools that could connect to a
laptop and be set up quickly - say on a routine customer site visit.
I've heard from several people that replays of people struggling to
use software have at least as much political value as value in
identifying usability problems.
I understand that if designers aren't part of your current
development approach, it's going to be hard to explain to folks what
there unknown unknowns are. And, I suspect others might have
diagnosed part of the issue correctly? Do you currently suffer from
lack of competition and large market share? Those are terrible
burdens on usability. ;-)
Thanks Jeff G. for posting!