Do we put too much blind faith in numbers as an industry? So how exactly can digital get real and measure more effectively for the sake of sustained growth?
'Not everything that can be counted counts, and not everything that counts can be counted' -- Albert Einstein
There are a number of measures we look at when assessing a digital campaign and marketing strategy. Some make more sense than others and some add no real value. Yet we still religiously track and report on them, ignoring other metrics which may be more significant. We often have blind faith in the numbers and the technology, showing a fundamental misunderstanding of what the numbers really mean.
Impressions and clicks
When advertisers buy ad space they more often than note purchase impressions or clicks. Measuring impressions and clicks is easy, but they are not the primary measure of success. Interaction rates, click stream and repeat visits -- as well as consumer interactions with all of your marketing components -- are what you should be measuring. Measure the entire track from exposure to success event (usually a purchase/acquisition) and your marketing efforts will be rewarded. Ensure that you know how much you spent and what increase or decrease you saw in sales -- that is the most fundamental strategic metric you can measure.
'But, my campaign's objective is to build my brand. So impression and click data is really all I need!'
Building a brand is not -- and never was -- the end goal of marketing. Building a brand is a transition metric to driving sales or increasing market share and that is the goal of all marketing. There is not one example of any marketer in the world that spends money with no expectations on ROI, so why should we pretend any differently? In addition, impressions and clicks are poor measures of your brand. The age-old measures of awareness, favourability and commitment are the right measures of your success.
Brand building is a good start into a relatively new metric that has marketers running off in the wrong direction: engagement measurement.
Engagement measurement cannot be described as the method of looking at the impact of all ad exposures and what contribution was made to the final sale. While Microsoft's engagement mapping -- or EMAP -- approach is a useful line of analysis, this is really attribution and there are significant problems with the approach:
- First, it allows advertisers and agencies to make discretionary assumptions about the expected impact of particular attributes (ad size, creative type, etc) in the absence of any statistical evidence that their assumptions are valid.
- Second, to partake in engagement mapping properly, you have to collect a massive amount of data and integrate all channels. Carrying out engagement mapping with only a few preceding exposures -- and without including all your channels -- will simply give you a partial view, as opposed to the bigger picture.
Engagement is really about measuring brand 'interaction, intimacy, involvement, influence', and ultimately commitment.
In today's Web 2.0 world, a visitor can engage with your brand outside of your website. The fact is, it's very difficult to understand total engagement. How do you know if a visitor is on your site, talking about you on Twitter and participating in a group on Facebook about your brand? Where does 'visitor engagement' start and stop?
Pageviews
Conceptually, pageviews seem straightforward, they can be described as a measure of the number of times users view your content. Sadly, it's not that simple. The internet is swarming with bots querying servers for content that no human will ever see. Furthermore, in the age of AJAX and widgets, it's hard to quantify what exactly a pageview is. Most importantly though, users are spending more time online, generating more and more pageviews. Facebook and Bebo are prime examples. None of this is news -- see last year's discussion;
Death of the Pageview -- but it's amazing how people still cling desperately to the pageview metric.
Customer data attributes These are often measured, but not used to segment databases. Demographic and geographic data are most often used -- they are easy and there is an umbilical cord to more traditional marketing and segmentation methodology. But what about using the way people interact with your brand, segmenting them based on this? Click and open email rates are seldom used when segmenting; what is their online spend? Where did they come from (acquisition source)? How long ago did the purchase anything from you? These are basics!
In conclusion, some metrics we religiously use add little or no value, or just don't make sense at all. We need to set realistic goals and ensure that we measure them with the right metrics. Unless this problem can be addressed, the inability to make accurate measurements of digital advertising's effectiveness will continue to promote the misallocation of marketing budgets and to impede the industry's growth.
Grant Keller is EMEA Director at Acceleration.