Editorial Articles

Over 100 years ago an American Retailer called John Wanamaker said: ‘I know that half my Advertising doesn’t work, the problem is that I don’t know which half’. Not a lot seems to have changed despite the revolution of the Internet, digital media and incredible advances in technology. I often hear Advertisers stating that they can’t get display advertising to work… The problem is NOT the media, its what is meant by ‘working’. 

Advertising is a form of communication intended to persuade an audience. Today’s de-facto standard of ‘last-click’ wins simply misses the point of what advertising is, how it works or for what tasks it’s been deployed. It’s essentially flawed on two fronts: The first is the ‘click’ part – it assumes that for an ad to work it must be clicked. The second is the ‘last’ part – it assumes that any previous impressions or touch point before the last one deliver no value.

The customer journey is often defined through the Sales Funnel concept which has established stages starting with Awareness, then Interest, followed by Desire and finally Action - AIDA. There are challenges to this traditional funnel concept when applied to the Internet and I’m not going to argue for or against one shape or another. I do however believe that there are stages; Initial consideration set, research phase, purchase etc. Therefore Advertising has to fulfil one or another persuasion task. For example, persuading the audience to consider the Brand or Product, persuade them to engage and ultimately to purchase.

Complex algorithm based Attribution works by identifying those ‘touch points’ that contributed towards a conversion event. Having determined which of all the touch points contributed, the correct partial value based on Recency and type of touch, click or impression etc, can be ascribed and the true ROAS or ROI identified. In short, we can address John Wanamaker ’s problem – we can identify the ‘half’ that worked. Having done this we can then analyse the relationship between the media channels, determine the performance of different placements and ad formats for example.

In a recent Case Study for one of the UK’s leading Cinema operators, we identified the value of display impressions. We did this by measuring site visitor propensity to convert based on the presence and Recency of display impressions. The Recency curve is not linear as you would expect but for the sake of a single defining number where there are recent impressions the propensity to convert increased by 36%. We then looked at the combination effect of display impressions on search clicks. Where there were prior display impressions before the search click the uplift in conversion was over 55% - quantifiable proof that ads don’t have to be clicked to deliver value.

I’m sure you will have read different numbers from different studies – often claimed as the generic uplift etc. Having carried many studies for different products and Brands – we can be categorical in stating that there is no such thing as a generic uplift. Every campaign is different and often for some obvious reasons. The maturity of the Brand, the extent to which the purchase is price led, utility led, impulse or planned all affect the period over which the value of the impression lasts…

Intuitively we know how advertising works – however we do need to adopt new measurement techniques to quantify them. Will your CFO sign off on a hunch or does he want real measures against which you can optimise your media? I’m convinced that the man recognised by some as the father of modern advertising, John Wanamaker , would have already dismissed last click and adopted complex, multi-model Attribution measurement.

Matthew Mills
CEO & Founder, Shomei