Pulling Back the Research Curtain

It’s been a crazy few weeks, with Thanksgiving followed by my move to Manhattan’s Upper East Side to jury duty today (and perhaps beyond), so I’ll keep this one quick. The latest Search Insider is basically a slap on the wrist to anyone (and particularly one repeat offender, Lycos) that tries to manipulate data to come off as something they’re not. Namely, Angelfire is not bigger than MySpace, despite what Lycos might have you believe. The full column continues in the extended entry.

IT’S NO SECRET THAT YOU can use research to support any case
you’re trying to make, but when the data’s released by a search engine, it
highlights some of the challenges in understanding search research.

The first rule of reading any sort of research is that if
something sounds offbase, it’s worth questioning. Maybe you’ll discover that
it’s truly counterintuitive research, and hence worth the extra time digesting
it. Maybe you’ll realize you just didn’t understand it, and once diving deeper,
it makes sense. Or, maybe the story you’re being told doesn’t hold water.

One announcement that required further investigating came from
Lycos, which reported last week that "its Angelfire site has once again claimed
the top spot as the number one community teen destination for teens ages 13-17
and tweens ages 18-24, according to the most recent comScore Media Metrix report
of Internet audience rankings for Oct. 2006."

I only have so much Web surfing time, especially for sites that
cater to the 13- to 24-year-old set, so I’m thrown for a surprise every so often
by certain sites’ popularity. Yet how could Angelfire be bigger than MySpace for
a certain age group? And, as a friend pointed out, how could it even be bigger
than Geocities, which was once synonymous with personal homepages and was, in
many ways, one of the first blogging platforms?

A review of other data sources suggested something was amiss.
Using Google
Trends, I plotted the search volume for MySpace, Geocities, Lycos, Tripod,
and Angelfire; this isn’t a perfect correlation for traffic, but it’s an
indicator. Since 2005, MySpace has dominated search volume, and the rest of the
terms are a blur, with Lycos slightly ahead of the others. When I removed
MySpace, Angelfire clearly ranked last. The same pattern emerged when looking at
Compete’s
site rankings.

I asked Lycos about some of the discrepancies, and a
representative wrote, "MySpace is categorized (as per comScore Media Metrix) as
a social networking site, whereas Angelfire is a teen community destination."
She later added, "ComScore considers Geocities (along with Tripod and Angelfire
on Lycos) as a web publishing/web hosting site." ComScore was contacted about
this story and had no comment on the near-final draft of this column sent to
them more than 24 hours before submitting it for publication.

The discrepancies discovered in the Lycos release and comScore
research typify a classic problem of search marketing. If there’s a disconnect
between the marketer’s nomenclature and popular consumer usage, then the whole
experience is futile. Marketers are from Mars; customers are from Venus. What’s
the difference between a teen community site and a social network? I never was
able to find out, but I’d argue that "social network" has emerged as the more
common vernacular.

The Angelfire story took even more interesting turns when
reviewing Hitwise data. Some of the sites trailing Angelfire in the comScore
teen community ranking include Quizilla (where members post quizzes, poems, and
stories), Neopets (a virtual pet game), and Tagged (which calls itself a social
networking destination, so how that differs from MySpace is beyond me).
Reviewing Hitwise data from February through November, Neopets has always been
the top-ranked site, Tagged and Angelfire have jockeyed for second place (Tagged
dipped in September and October but rose in November), and Quizilla has
consistently ranked last.

In another snapshot of the data, I was able to review a Hitwise
grouping for "Web 1.0" sites–namely, Geocities, Angelfire, Tripod, and WebRing
(listed in order of market share). Collectively, their market share of Web site
visits plummeted from 0.33% in November 2004 to 0.14% two years later. Even if
raw traffic numbers have increased for any of these sites, the category overall
doesn’t seem to be one of the high growth areas for the Web. There were other
signs something’s amiss with the claims in the Lycos release. It mentions that
the five most popular topics with Angelfire members in 2006 include do it
yourself/how-to tips (#2) and addiction recovery (#3). Maybe, somehow,
Angelfire’s teen users are big how-to junkies (how to illegally download music,
how to find your nearest H&M, how to score a date with members of Congress).
Even so, there’s no way that addiction recovery makes it to third place for
teens, topping sexuality and music and just two notches down from relationships.
Then again, perhaps that included addiction to Wii and PS3.

Some day, I wouldn’t be surprised to see Ask.com issue a release
that it’s the No. 1 search engine, since Google, Yahoo, MSN, and AOL are all
portals. The trick is to use the numbers well enough that they support your case
and boost your standing without drawing attention to themselves. If they draw
too much attention, someone’s going to come along, pull back the curtain, and
see if there’s a real wizard there.