[ 2] Armano, D. “How
Filter Failure Contributes
to Business Failure”;
http://darmano.type-
pad.com/logic_emo-
tion/2009/09/filterfailure.
html#comment-
6a00d8341bfa9853e-
f0120a5eeb6e6970c/
March + April 2010
interactions
are talking about in regards
to their brand or domain.
The current, most prevalent
tool for mining these digital conversations is sentiment analysis,
which allows anyone to search
for keywords and sentiment, i.e.,
“____ sucks,” “____ rocks,” “I love
____ brand,” “I hate ____ brand.”
At this basic level, you don’t get
much other than the whiff of a
good or bad vibe related to your
brand. A recent New York Times
article makes mention of lightweight tools [ 2]; we think of them
as experiments in data mining,
a less useful but more interesting way to get a sense of general
customer sentiment. More powerful tools trawl the Web from
thousands of sources—as many
as they can tap into. Some companies (like Zappos) incorporate
casual data from listening by
consolidating it and allowing it
to be explored by their UX team.
What it’s good for:
Monitoring the effect of a •
new product launch or ad cam-
paign
Gaining competitive •
intelligence
Listening in on pain points •
Developing ideas for future •
Gaining inspiration from •
consumer suggestions (design
ideas)
How:
Utilize existing tools and •
services
Leverage unique key words •
Analyze for sentiment and •
Map to related timeline •
(against events, trends, etc.)
Try this out: Go to search.twit-
ter.com,type in a brand name—
Nokia, maybe Wendy’s, anything
you want—and see what you
can find. Listen in. Add some
sentiment queries on top, get
creative—“____ sucks” or “_____
awesome.”
The Challenges: Searches are
key word-specific, and most data
comes without strong demo-
graphic or user information to
help in synthesis. Getting rich
meaning from this data is dif-
ficult.
3. Engaging; reach out to the
customer directly and have a conversation. Establishing a method
of collecting and acting on consumer issues faster, and nurturing relationships with users to
help build and evaluate products.
We see two methods of
engaging: Acting on listening
quickly in order to limit the vocalization of a bad experience and
deliver resolution to the customer. At Zappos; we bought shoes
from them; they came in the
wrong size; we complained on
Twitter; Zappos responded to us
instantaneously, asking how to
make it right. It utilizes the Web
to pick off bad-sentiment scenarios like this and deliver good
customer service in real time.
Revealing the bigger plan. Think
of this like co-creation techniques in research: You bring
the customer into the design
process, make them part of
your team, and tell them how
they will matter. Starbucks’s
MyStarBucksIdea is the best
example of this kind of engagement as it exists online. The
company crowdsources its
customers for ideas, lets them
vote on the ideas in a Digg-like
fashion, then pushes high-rated
ideas through the innovation
process and shares the project
timeline so that contributors
(customers) get a sense that
their idea is being acted upon
and can track it. Starbucks took
farming to the next level and
communicated the outcomes of
user input back to customers.
What it’s good for:
Immediate customer service •
(nip something in the bud)
Creating listening posts for •
consumers to engage
Employing farming methods •
to get feedback
