[Sigia-l] Jared Spool's article on galleries

[Listera]

Olly Wright:

> I can try. Something like "How do you enable user-friendly product
> selection when the products are complex, and users both have little
> domain knowledge and do not want to have to acquire much specific
> domain knowledge?"

Great. So would it be fair to say that the fundamental problem is *not*
about "galleries" as described in Jared's article?

Would it also be fair to say that the fundamental *design* problem here is
the complexity of the product and that as long as the product itself remains
so complex (wrt user expectations as you describe) then specific model
selection will most likely continue to be problematic in so far as it tries
to mask the inherent complexity? (In other words, this is not how, say,
Apple would solve this *design* problem. :-)

So could we generically reformulate the problem (not as redesigning
complexity) but as camouflaging it?

The first principle of camouflage is that the object appears to conform to
the context. However, once you set up the context (i.e. a picture gallery)
it's very hard to switch to another visual context/paradigm. So it's best to
match the context and the user expectation upfront. No pogo sticking
required.

If you observe people trying to describe what kind of a product they want in
a store, especially when they don't quite know what they want, you'll see
that they try to verbally describe it, but if they can show or, better yet,
hold a similar product in their hands, they then proceed to add/subtract
features from it: size, color, weight, buttons, etc. with words and
gestures.

In other words, it's brain-dead to ask people to extract commonalities or
differences from an array of products they know little about. It's much
easier to ask them to describe or, better yet, show what they want in a
product and then do the mapping *yourself* to the product line.

So the approach would be not picking but describing -- a subtle but vital
difference.  

I did a Flash-based make-up studio design for a French company a couple of
years ago where the user essentially painted a face by clicking on various
make-up options. A visual product list, mapping to what was on the model's
face at any given time, would dynamically reconfigure as make-up changes
were made. You could optionally present the user with "defaults" to start
with by letting them select from a photographic list of model faces. This
works for cars, shoes, clothing, etc. When well executed, works great.

One recommendation I'd also make is to ask the user what phone they are
currently using, if any. You can then immediately map out relevant products
from your line. That's a great starting point, "default." You can then ask
the user what they do/don't like about it, and how they'd like to make it
better suited to their needs. If you have a phone that can fulfill that,
you've probably made a sale.

So to sum up (man, I haven't even had my eggnog yet :-) don't start with
what you have but with what the user wants.

----
Ziya

"Innovate as a last resort."