Category Archives: Quantcast

Gmail goes down


Gmail’s usage goes down, that is. We’ve been reading recently about shifts in consumer media habits, some surprising (don’t miss Wired magazine co-founder Kevin Kelly’s analysis showing TV is the fastest-growing medium in the U.S., still far outpacing the Internet). And then we thought, we don’t check our personal email that often anymore, since we rely on Twitter and Facebook and messaging for intimate communications.

Here’s what Quantcast reports, based on direct cookie observations of traffic to the Gmail site. Down 50% in one year. It’s an amazing trend, and one to watch as Facebook enters the email space with its Messages.

Rapleaf maps human relationships for offline sales


One of the coolest behavioral targeting trends we see is from emerging companies such as Rapleaf that map the connections or relationships between consumers. This is revolutionary ground, because companies can now predict what you might do based on what your closest human friends are doing.

Consider that a recent study by AT&T found that “network neighbors,” or people who communicate often with each other, are 5 times more likely to respond to the same marketing offer, and you see why marketers are intrigued.

Here’s one example of how Rapleaf enables this. Imagine you run an airline company and map data to realize that customer John Smith flies twice a year from New York City to Las Vegas. Now imagine you append information from Rapleaf, compiled from millions of online social interactions, and find out that John Smith has two college friends who live in Las Vegas. Putting two and two together, you could then send direct mail offers to John’s friends in Vegas offering them discount airfares to visit New York City — so they can fly to see their friend in reverse.

What’s interesting about this?

1. It shows the power of social media in mapping real human connections that can make offers more compelling to consumers.

2. It shows that social media is a data collection source — but the marketing implementation could be somewhere entirely else. If you can learn something about consumers by watching their behavior on Facebook or Twitter, the best sales approach may be with traditional offline marketing media.

3. It shows that people are watching you. Privacy guidelines aside, now with so much exposed online, for the first time the personal relationships and individual quirks are going into master prospect databases.

Photo: Josef Stuefer

Seeing the who, not the what: Robert Scoble’s interview with Quantcast

Most web analytics measure what happens online — how many impressions, clicks, conversions. But finding out who is visiting sites has been more difficult.

We just chatted with Konrad Feldman, co-founder and CEO of Quantcast, for a forthcoming BusinessWeek.com column. We won’t spoil the thesis, but if you are interested in web measurement, the Quantcast system is free and provides extensive demographic profiles of web site audiences. Here Robert Scoble interviews Konrad for details of how it works.

Meet the Googlesexual


Who knew IT could be so studly?

A colleague coined the term Googlesexual as a joke this week as we were researching gadget sales among metrosexuals (hip heterosexual men fascinated by clothes and hair products, say, Tom Cruise) and übersexuals (similarly hip men who focus less on mirrors and more on causes, perhaps Ewan McGregor). Neither term is derogatory, but instead pinpoints a state of mind in how people view themselves — cool, stylish, cutting-edge — which in turn connects with their behavior responding to advertising.

The more we considered it, there is a new species of psychographic variables (interests, attitudes, opinions) best represented by Larry Page and Sergey Brin. Googlesexual. The “sex” doesn’t mean procreation, but rather the deep Id inside the soul that drives instinctual behavior. The “Google” doesn’t mean a single company, but rather the power that comes from understanding networked technology. Put them together and you get the IT guys now running the planet, the manly risk-takers behind Internet Bubble 2.0, the women setting up media empires with blogs, Clark Kent with an attitude. The world has always had “dandies” and “dorks” — but what about men and women of technological sophistication who wield power and affluence and aren’t afraid to blow VC funding? These wundergeeks need a new name.

This is important, because Googlesexuals are driving the early adoption of most technology and entertainment. It’s not about age, or income, or Gen Y or X. It’s about how you see yourself, and how that invisible mirror spurs your reaction to products.

Many marketers miss thinking about psychographics and instead use baser levels of planning:

0 – base level – all about the product. “We made this widget, and it’s great!” This type of marketer wants to buy advertising as cheaply as possible, then spray and pray.
1 – media level – all about the advertising channel. “We want to buy ads! On Lifetime!” This marketer usually calls asking for a specific TV cable channel.
2 – demo level – about the customer demographic. “Hmm, after deep analysis, we understand we are targeting men 35-44 with household incomes greater than $100,000 in urban perimeters.” This is the most common marketer; armed with research, he or she knows where customers come from and what exactly they look like.
3 – psychographic – about the human mind. “Our target customer has these attitudinal aspects toward her self-perception, sexuality, friends, products and media. She is a Googlesexual who prefers to tell her friends first about technology trends, share advice, and networks with more than 1,000 people online.” Marketers who plan at this level have the best odds of using media to influence the marketplace.

Ideally, a psychographic target allows you to pick the media and message most likely to resonate. But it’s tough to reach Googlesexuals, because online they are hidden. At the OMMA conference in New York City last month, we witnessed managers from comScore (which uses projections from small groups of consumers in research panels) and Quantcast (which uses cookies) get into a heated argument over whose methodology was best at tracking demographics of online users. Consensus was, most web site demo stats are bogus, since it’s still impossible to really tell who is typing at the keyboard.

Difficult to find, but not impossible. IPhone sales tell you they are out there. Amrita Chandra, a Toronto friend who runs an art gallery, had perhaps the best response to our Googlesexuals vs. übersexuals thesis: “I’ll take one of each, thanks.”