• by David Kang , January 4, 2008

I AM THRILLED that Web analytics is growing up, that standards are increasing and more people understand there are ways to use data to make better decisions.

The one thing that I am not so sure about is the implication that Web metrics can adequately measure for brand impact, awareness and other measures traditionally used to approximate value from upper funnel marketing activity. I have seen many stop here and say that Web metrics do adequately account for value (such as sales, data capture, or loyalty), and if awareness does not correlate in that analysis, perhaps the value (or lack thereof) of awareness is to blame. I think this an interesting position and worthy of debate, and I promise to devote future space to discuss that issue in detail.

 Getting back to the topic, take the simple metric of clicks. Assume a common situation where Web-based ads drive to some action (typically, drive to a site). Many marketers still use clicks as a directional measure for awareness (the simplified argument being that one could not click without an awareness shift taking place). But there are challenges. Given that 99% or more of consumers generally do not click on an ad, a directional understanding of the less than 1% that do may lead to a sampling bias. Of those that click, not every click is equal. If you have mined data from the perspective of a cookie, you know that a click can lead to a 5-second engagement or a 5-minute engagement. Some convert to sales, some convert to email address captured, some go no further than time spent. Some lead to enormously loyal customer behavior, some seem frankly accidental.

 What other Web metrics can we examine? Would impressions, or time spent on an ad unit, any type of conversion rates, provide better commentary on whether or not consumers have become more aware of your product as a result of your advertising activity? I think they all add interesting commentary -- but opportunity exists for improvement.

We can swing over to the use of surveys as one quick answer, and for the moment, I find it hard to argue against this. (Except for the normal limitations present with surveys such as biases introduced based on how the questions are written, etc).

Here is what I think interesting to advance the measure of awareness.

1) The rise of coordinated Web metric and survey tools. Deploy a full scale Web campaign, so you arrive at the usual complement of metrics. Deploy a survey (at the ad unit or web site level), and utilize cookies to marry the two sets of data together on a consumer level. Taken on a large scale to defeat sampling bias, one can start answering questions like "did consumers who say they became more aware actually provide more value through their Web actions?" Tools with prevalent cookie reach can probably benefit greatly from this approach. Kept within a marketer's walls, and kept anonymous, there should be little legal concerns.

2) Word of mouth and buzz tracking. I am getting increasingly more interested in word-of-mouth tracking. It can give you objective data like you get from Web metrics, but also qualitative input not biased by the structure of a formal survey. Coordinating word-of-mouth tracking with a web metrics tool (and maybe a survey) might also be interesting as a way to secure full sales funnel visibility.

I invite you to share your comments. How have you wrestled with arriving at better visibility for awareness and reconciling awareness and response objectives