The Warc Blog

The Warc Blog

Pretesting, BE and neuroscience
Manfred Mareck, Managing Director, Research Marketing
Manfred Mareck

Apart from interesting presentations, I find advertising conferences have another benefit: catching up on commercials that have otherwise completely passed me by. Sadly, the range of worthwhile examples seems rather narrow: a couple of years ago, no advertising conference would have been complete without numerous references to creative masterpieces from Lynx/Axe, Dove or Cadbury. All these clocked up millions of views on YouTube which, I guess, makes for a successful campaign.

The likely success of such campaigns can be significantly improved by rigorous pre-testing. This was the message from Peter Haslett (chairman of Ipsos-ASI) during the session on Tools and Techniques. Ipsos' pre-testing results, said Haslett, correlate highly with actual sales and CEP scores (cognitive emotional power) which help identify whether emotional or cognitive/rational forces drive impact. As is often the case, brand advertising that combines the right mix of emotional as well as cognitive power work best, but usually the purely emotive ads are the ones winning awards.

What's important when pre-testing is to understand the (big) idea behind the campaign, measure what really matters, factor in today's complex media environment and validate results.

Behavioural Economics

Dr Nick Southgate started his presentation on Behavioural Economics (BE) by saying that the success of this discipline in stimulating the creative industries is because behavioural economics is completely disinterested about those things. But it clearly has some value, as the IPA appointed him earlier this year as their BE consultant.

BE is an attempt to build a theory of economics out of principles derived from observed, rather than postulated, human behaviour and to go beyond the axioms of classical economics, such as rationality, preference, self-interest and so on. BE uses obliquity, i.e. the analysis of the most unexpected things.

Southgate uses the example of ordering pizza: one option is to build a pizza up from a bare base by adding ingredients, the other is to be offered a pizza 'fully loaded' and the customer removes unwanted ingredients. People 'choose' more ingredients in the second condition - clearly the promise of a mushroom you have to 'give up' is apparently more delicious than a mushroom you add.

BE asks this question whereas classic economic thinking simple assumes that preference is not distorted by the way you order a product. BE can teach advertisers that an offer of say 10% discount on a product will result in fewer units sold than an offer that says '10% off - maximum 5 tins/packs per purchase' - value, it seems, is more meaningful to shoppers when the offer is limited.

Last week, I reported on the irresistibly titled Chief Happiness Officer Stephen Phillips (Spring Research) from the annual ESOMAR Congress. For Warc, his presentation focused on Emotions and Engagement. Emotions, said Phillips, guide most of our decisions and the marketing industry has somewhat neglected emotions and many of its models relied for too long on linear, AIDA-type concepts. Conjoint analysis also typically assumes too much rational decision making. Purchase decisions are driven by a complex interplay between emotional impulses and rational/practical considerations. TV often works best at the emotional level; print can be stronger to influence rational decisions - thus different media channels should be selected for each phase of the buying process.


Graham Page (head of innovations at Millward Brown) opened with some caveats: whilst neuroscience measures biological responses (thus overcoming the bias of direct questions), there are practical issues (size and cost of the scanner) and issues with the current hype (fMRI is provides a better, but still not full understanding of how the brain works). In other words, it is not replacing but enhancing traditional methods. Whether insights derived from neuro-science or biometric measurements are ultimately better predictors of human behaviour is also not clear at this stage, as there is simply not enough validation available as yet. Page then showcased three key methods/techniques:

1. Implicit association measurement (measuring reaction time/accuracy to ideas or concepts to infer emotional responses - for example, what ideas brands evoke in respondents).

2.Eye tracking (provides very detailed insight why individuals respond to ads, how they orientate themselves in a retail environment etc). Eye tracking results for advertisements, for example, show that, when asked, people tend to overestimate the attention they pay to brand/logo/strap lines etc.

3.EEG (brainwave measurement) - for ad/creative diagnosis. Uses a head cap (rather than full MRI scanners) to measure positive/negative feelings and establish whether respondents feel cognitively involved with what they see.

Dr Carl Marci (CEO of Innerscope Research) urged researchers that they need to dig deeper to measure where responses actually happen. Conscious responses represent only a small portion (somewhere between 5-25%) of what the brain is doing, while unconscious measures go beyond conscious measures by measuring emotions, which direct our attention, enhance learning and memory and ultimately influence our behavior. Biometrics and eye tracking are an important complement to existing research tools and a natural evolution of a movement toward ever higher sample rates and granularity of experience.

He sounded some caveats: EEG only measures brain wave activity within the outer part of the brain, i.e. it does not reach the deep brain structures where emotions are processed. Eye tracking can only show where people look, and is only of value if it also measures the points on which the eye fixates during the saccadic moves to provide not just conventional heat maps but also gaze plots. Measuring emotions is complex and requires multiple channels but when combined can provide a comprehensive picture of our emotional engagement.

Subjects: Warc conferences, Digital, Advertising

24 September 2010 12:30

There are no comments on this blog

Comments IconAdd your comment here:

Blog Search


  • 2016
    • October (13)
    • September (11)
    • August (18)
    • July (11)
    • June (16)
    • May (10)
    • April (17)
    • March (16)
    • February (13)
    • January (12)
  • 2015
  • 2014
  • 2013
  • 2012
  • 2011
  • 2010