How to write an awards paper

John Grant has played devils advocate and given a great demonstration of how to write an awards paper on Adliterate.

“I had a funny idea: why dont I try to ‘prove’ to you that the Honda UK advertising was extraordinarily successful?
(I think this strand should be renamed ‘devils advocate’)

I think the argument would go as follows;
1. the car market is very change resistant for four reasons; long purchase cycles (4+ years), cognitive dissonance (current owners being indifferent to any communications that dont confirm their current choice), loyalty (60% buy the same model again), historical sales bias (the huge second hand market which feeds the new new car market with loyal buyers and also determines what you see on the road/regard as ‘normal’). Prof Ehrenberg concluded his classic study of the car market with “there are no strong and weak brands, only large and small ones”
2. over a longer run all marques introduce new models and the efffect of this in any one year (being more ‘in the news’) evens out. Only “star’ models of the sorts which win awards make any lasting difference. Honda has one of these in the new Civic, but it wasnt released until 2006. In the period 2001-05 a quick google revelaed no major star models (ie UK awards)
3. Because of the long purchase cycle (and 60% loyalty and indifference), a gain in share one year will not be carried over into the next year (not until the increased base of buyers come to rebuy – in Honda owners case I bet their cycle is longer than most, but lets say at least 4 years). If you had a great ad which produced 10% higher consideration and purchasing in 2001, in 2002 your natural base of sales would not start 10% higher. You would be starting from scratch. So the data showing jumps of 10% a year are really showing jumps of 10%, then 20%, then 30%.
4. only if you had a cumulatively effective campaign that actually did cut through indifference and persuade more owners of other cars to put the brand on their shopping list, because it permanently changed the image of the marque, would the gains would be sustained. this would be an extraordinary achievement there are no other known examples (eg effectiveness awards) for doing this except in the case of Skoda and arguably here it was a case of VW changing the brand, not just adverts.
5. The way to measure effectiveness is to look at the increase across a purchasing cohort of say 4 years; in this time people will not on average have repeat purchased and there is time to shift brand perceptions (most reckon 3 years+); you should look at the overall sales increase, and also the consideration scores.

In Honda’s case these are +55% and +100% (very roughly)

You need to show that no other gradual factor could account for the same phenomenon; eg the pricing relative to main competitors (exchange rates), the number of dealer outlets etc. But it is much easy to eliminate most other causes like new star models or competitors having a crisis because they would produce step changes in fortunes, not gradually rising sales. Only changing the consideration pool can do this.

Data from other markets show that Honda is also doing very well in some regions with new dealer networks (Sweden, China) and is flat within mature markets (Japan/USA). But there are no cases (I assume) where it is established but weak (as in the UK) and has grown substantially.

Therefore this is an extraordinary success”