[Sigia-l] Why Good Content Must Suck: Designing for the Scent of Information - Jared Spool

[Derek R]

__________________________________________

Definition: "Information Scent" -- The *subjective* sense of value and
cost of accessing a page based on perceptual cues; Scanning; Perception,
being, arrangement.

'Information Scent' seeks to answer questions concerning the interests
of visitors to a Web Site, and predict their paths through the Web Site
*given their interests.*

So that -- 

Result #1 -> Given a user information-need, it is possible to describe
an algorithm which simulates usage of the Web. 

Result #2 -> Given a user path through the Web, it is possible to
describe an algorithm which infers the information needs expressed by
that path.
__________________________________________


The most important thing to note from this 'scent/foraging/scanning'
method is that Result #1 proceeds from a *particular affirmative
identity* (user-interest=true=X=known). Therefore, YES, this is a good
method because it uses good reasoning (X = X + 1) and does not attempt
to build its case the opposite-way, which is false (X + 1 = X) since the
assignment is unclear.

The next most important thing to note is that Result #2 proposes 'needs'
(plural) and seeks to present *all* the possible/many 'needs' which
could be ascribed to a 'singular,' known, path (again X = X + 1, ok).
However -- *which particular* need (singular/truth) we are actually
talking about (X=the user) cannot be ascertained from Result #2 unless
Result #1 is *also* known! (i.e. as in Result #2, to know which
'information need' is expressed by a known path we need to be 'given'
the particular/actual user information-need, otherwise, we can only
supply many/possible 'needs' (plural). In this sense, Result #2 is
*useless* in so far as it leaves us a diffuse everywhere, which is
nowhere in particular -- it is not definitive and is akin to building
your house on the sand).

This is the problem I have with ideas which promote a *particular* path
(singular) or 'scent' as inferring a *particular* need (singular).

In short -- the only way to know a user information-need is to known
that need, specifically. There is no golden ticket, or short-cut,
despite marketers efforts to *lead* you to that conclusion. The user
*must* have to choose themselves (particularly). This is the only way
which is genuine, capable, and real. 


Summarily: the *best practice* is plain and simple 'derivation.' Nothing
fancy in that word. It is not trying to 'sell' you anything, only
presenting the truth -- that the more specific an idea is the more
distinct it is, and consequently, the clearer it is. 

In other words, if we know *what it is* a user wants, we can give it to
them, it being known. If we do not know what a user wants, just giving
them something (consumerism) can NOT presume to have fulfilled the need.

[ Consequently, 'findability' works this opposite way (which is false)
by proposing that 'if we can give it to them (make it findable) it is
what they want' -- an end, marketing. I will continue to expose this
falsity until the people who purport to call themselves IAs drop this
term, and its meaning, from their vocabularies, or show/prove me to be
wrong. Presuming something is absolutely useful to a person just because
it is available/for-sale is *marketing* and not information
architecture. ]