Its all about sex .. and if people think it will help them get it (or even the hope) they will follow the ad to the shops (and beyond).

The “super models” are a great pull for men .. but most women will be looking for the flaws in the shot ! “Oh her skin is awful” … men will not normally notice blemishes (and air brushing is SO forgiving!) But women will look for the imperfections to justify themselves that THEY are OK too ! .. hence the short glance at the super models because they know them .. and know they have been “fixed” in the darkroom/ computer .. they know these models are not real .. but men fall for it every time !!

Quant online pre-testing is especially valuable to category captain such Tide and Tropicana, indeed I’m interested to know if the Tropicana fiasco was pre-tested as I expect it was not: If it had been it *SHOULD* have been weeded out and saved shareholders $100’s millions of lost value. This is where Millward Brown reigns supreme; Link testing is the best solution for circumstances where the downside is enormous and the upside is small.

I believe quant online pre-testing is less useful for brands that are failing, in danger of being delisted and in need a big idea, such as Old Spice and Hovis. But these cases are unusual. There are many more brands that are simply treading water, Cadbury’s Dairy Milk for example. Sales were flat lining in in 2007, there was an over reliance on price promotion and a big pressure on price. As we all know the solution was Gorilla that delivered an ROI of £4.88 for every £1 spend. However It took six months to convince internal stakeholders to run the campaign, this wasn’t helped by the fact that the ad scored low on Millward Brown’s Link test branding score. When it was released the ad achieved the highest ever branded recall ever with a score of 96%. The IPA have the case study here: http://vimeo.com/31794837

Millward Brown has a great position in the market as risk mitigators for many of the world’s best brands, but Link is not the best solution for creative work that challenges the boundaries of advertising: Millward Brown had ample opportunity to put the case for Link testing supporting creative ads in an article in an article in Admap Nov 2011. Instead of making a compelling case (when given editorial free reign) an interpretation of the data presented is that Award Winning Adverts are Irrelevant and Incomprehensible?

If you are Tide in the US, have 70% of the detergent market, and plan to outspend your competition 3:1 then a reasonably low-risk approach makes sense. Your potential upside is fairly small and your potential downside is huge.

If you are Old Spice, the opposite applies. You may disappear or be delisted unless you do something remarkable. On the other hand the possible upside is enormous. There you should be far braver.

The other approach is the barbell approach Taleb recommends. If you are Tide, spend 80% of your money very conservatively and experiment with 20% of it on higher risk ideas……