Demographics for clicks, clues for offline behaviour?

While browsing through the headlines a while ago, my eyes fell on a story from eMarketer about Online Ad Clicker Demographics. While only surveying US netizens, they've found that age, income and visit frequency are closely related tousers’ likelihood to click on ads. The article refers to an August 2008 study by iPerceptions, a web analytics firm. I've evaluated more than a few campaigns and websites where I've been very curious to know who hides behind IP addresses and mouseclicks, so these are very interesting numbers to me.

It points out one of two different directions in current analytics for websites and banner ads, where the other is preached by companies such as ROI Solutions (of Tribal DDB) whose blog has an interesting article with "A Call for Using Only Financial Metrics to Measure and Evaluate the Performance of For-Profit CMOs". This thinking is supported by others as well, such as the planner Nina Åkestam in an article in Resumé a while ago, arguing that for any campaign, the only thing worth measuring is profitability. I don't suggest that the metrics for evaluating the people in marketing and the ads should be the same, but the trend is cleary a call for metrics for what's really in your wallet.

While both of these directions are very interesting, are we missing something important here?

I'm assuming that the holy grail for all my analytics friends would be to track their users across all digital media throughout their day, such as corporate websites, banner ads, widgets, email, mobile and all the rest. The data keeps piling up for each new medium, but I rarely see it turned into real insights of how different people behave.

I want to put out a call to bring social knowlegde into web analytics. Let's make an effort to unify what knowlegde the media buying agencies have about user trends and demographics, combine that with behavioural patterns emerging from search and web analytics and voila - we'll have some pretty powerful insights about how different users behave online and communicate.

For example, if you are IKEA and you learn from your web analyst team that people from the west coast tend to spend a lot of time looking for new kitchens, while the northerners are really more into wardrobes, how about setting up your DM department to tailor for these needs? Or maybe you'll learn that traffic from your ads in an upmarket online magazine are using the search function on your website to look for kitchen appliances, and mobile users mainly look for childrens furniture? Should that change how you make your print adverts?

My idea, and I guess it's pretty obvious, is that what we learn from our consumers online tells us a whole lot about what they except from our offline communication. Let's fill the gap between what the analyst evangelists are eagerly giving us and the traditional social insights that account planners dig their heads into. The first one to be able to combine these with automatic tools is going to have a serious advantage.

And that's really what will bring in the money.

I want new metrics!

Some time ago I read an article about a study performed by comScore with Tagoda and Starcom which documented that within the U.S market, 6% of the online population account for 50% of the clicks on display ads. Not a surprise that clicks are not balanced across all online users, really, since we all behave and respond differently to advertising in general.

However, problems for those of us who are not dedicated number crunchers do not stop there. I was looking at the weekly traffic statistics for the major Norwegian websites the other day, and noticed that the most visited site (www.vg.no) boasts a whopping three million "unique visitors" in one week. Given that we are a modest 4.8 million inhabitants in this country, and 84% uf us are online (SSB Statistics, Q2 2008), this means that VG.no should have an impressive 75% of us visiting their website if "unique visitor" was corresponding to an actual person. It may be because I'm not a keen reader of VG.no, but that does sound a bit inflated.

With a recession looming ahead of us it's fairly obvoius that marketers are going to be giving us a clear message about kindly giving them the utterly best return on their investments. For those of us who work with interactive media, we need to get some new tools to be able to compare not only how different online media perform (how much is a 'click' on a banner worth compared to a 'unique visitor' on a website? What is 12 pageviews compared to 12 minutes in an engaging Flash game?), but also how they compare to offline media.

Microsofts analytics department, The Atlas Institute, takes a plunge into the deep end and makes an attempt with "Engagement Mapping", which may indeed be taking us a few steps further.

Now that Joost is back in the game with their new in-browser player to compete with Hulu, I'm very interested to see if traditional TV metrics will be moving towards a new standard as well as internet metrics finding new ways to prove efficiency and effect. Not to mention Google TV Ads and the new overlay ads that are popping up on nearly every web TV player I've seen the last few months - what will they add to this conundrum?

Exciting times are ahead in this sea of numbers, but give me something more than clicks and unique visitors - I want effectiveness!