: So you've got this flood of data and
you're filtering it down and further down and further down to
figure out the core -- where the memes, the conversational threads,
really gain traction.
Ghuneim: Exactly. You know, it's never been
about just having the data -- there's too much data, right? It's
understanding how not to abuse it, how not to get run over by it.
Like, you could behavioral-target with, like, a scalpel and a laser
right now, but does it need to be that incisive? If you
really understanding who's most influential around a subject, then
you can begin to optimize messaging; like, you can figure out the
best way to target future advertisements. It's also about knowing
when to message -- when to listen and when to talk.
Dumenco: Talk to me about listening to
negative talk as a marketer. Because while a lot of brands
are thrilled to engage in conversation with quote-unquote
fans, they can freeze up when non-fans start speaking up
on Twitter, Facebook, etc.
Ghuneim: You know, I think everything I learned
about social media happened before social media. Looking back to my
days of what I did before Wiredset, I remember one particular
situation that helped me hone my crisis-management thinking. I put
up one of the first rock-and-roll bulletin boards at the time.
Which tells you that I'm old, but, you know. [laughs]
Dumenco: When was this?
Ghuneim: It was probably '94, '95. I was
working with Toad the Wet Sprocket, the band. Toad wanted to really
engage with their fan base, so the net comes about, they understand
the importance, we're doing some really fun stuff with them, and we
get to deploy the first bulletin-board software. We put up a forum
for Toad the Wet Sprocket, and the first post is, "You guys totally
suck." I get a call from the manager and the head of the label and
the product manager and the head of the fan club and they're all,
like, "Take it down, take it down, take it down, oh my God, take it
down!" And I said, "OK, OK, OK, OK," hung up the phone, absolutely
didn't take anything down, and then sat back and watched the band's
fans trash that post down to dirt. They made that guy go away and
in doing so really started to form the basis of a community.
Now we're in the same place. I'm a brand -- did that person just
say something that's really damaging to my brand? Well, same thing:
Should I wait and sit back and see what's going to happen, or
should I act immediately? These same decisions are being made now.
Sometimes you don't want your first decision to be an act of
censorship, because from there the second and third decisions and
so on can just get worse and worse and worse. Rapid-response brand
management today is about understanding how to communicate and how
to give your brand advocates an opportunity vs. worrying about the
detractors all the time.
Dumenco: It strikes me that you're in a great
position working with entertainment and broadcast people, because
they've understood for decades a kind of almost slow-motion
real-time marketing. Like, a label would watch their artist perform
on "Saturday Night Live" and then breathlessly wait for the
Billboard charts -- and the retail-scan numbers, when SoundScan
came about. Then it became about even more instantaneous metrics:
site visits, MySpace song streams, iTunes downloads, Amazon sales
rank, Twitter buzz, etc.
Ghuneim: The window has just gotten smaller.
That's where Trendrr came from -- we needed more fine-tuned tools
to understand what was happening in the marketplace to communicate
to our clients.
Dumenco: And then Trendrr took on a life of its
Ghuneim: Yeah. I wanted to be an agency that
builds things -- that makes beautiful things, ultimately. We
actually started Trendrr, which we called Infofilter at first, with
two data sources: Delicious and Flickr. We took a big, big risk in
building out a product at that time and continuing to build it
assuming that the net would open up and data would open up, and
sure enough, it did. Delicious and Flickr were among the first to
open up, but then slowly things started to kick into place.
Dumenco: What would you learn from Flickr?
Would it be people posting pictures that they took at concerts for
band"brands" you were monitoring?
Ghuneim: It showed how active the audience was
so, yes, you could look the next day and see events driving
actions: the taking of a photo, the conversation around that photo.
It was a brand-imaging tool: This is what's said about your brand
now. You might have the most sophisticated branding
strategies in the world, but if people do a Google search or a
Flickr search and see a particular conversation about your brand,
well, in that moment, that's what your branding is!
Dumenco: And now you've got dozens of data
sources feeding Trendrr. Which reminds me why I'm talking to you in
the first place: the relaunch of Trendrr, aka Trendrr v3. Beyond
the fact that Trendrr now also incorporates additional feeds like
location-based data from Foursquare and Gowalla, what's
Ghuneim: The streamlined workflow and feature
selection of Trendrr v3 is what sets it apart from the previous
version. Trendrr enables marketers to listen, measure and respond
to the conversation about a brand, service or product. The "listen"
layer enables users to search, curate and save real-time
conversation streams. The "measure" layer enables users to measure
conversation and activity across Twitter, blogs, news, social
networks, search, sales and location-based services, addressing a
shift to what we call "swarm consumer behavior." The "response"
layer enables users to communicate and market directly from their
Dumenco: You've added Facebook data too,
Ghuneim: Yeah, Facebook "likes" and shares.
Dumenco: I know some marketers and media people
would be blissed-out to have all that data at their fingertips, and
others would freak out at the thought of it, as in, Not more
data! I'm already swimming in too much data!
Ghuneim: The point of Trendrr is to structure
and contextualize large, otherwise incomprehensible data sets in
real-time by organizing sources into individual management
dashboards. Each dashboard has been developed and designed to
provide the user with simple, actionable intelligence. Our data is
actionable because it can be viewed in both real-time, as it's
processed, or over time, which provides a historical
Dumenco: Do you ever get sick of any of the
data yourself? Like, have you discarded any data streams because
you figured out that they really don't tell you that much?
Ghuneim: Yeah, you can get too mired in the
not-important ones. Like, yes, I can count the number of comments
on a video on You Tube. How actionable that is right now? Not
Dumenco: I think it probably just means too
many people have too much time on their hands.
Ghuneim: [Laughs] We don't hold it against
~ ~ ~
Simon Dumenco is the "Media Guy" media columnist for
Advertising Age. You can follow him on Twitter @simondumenco.