The problem with endless over-optimisation is that practitioners look at advertising through the lens of a spreadsheet which, argues Faris Yakob, loses the humanity in the craft.
This week I spoke at a lovely event in Belgium for performance marketers, held in a massive car museum. The theme was conversions – as in increasing them in your digital advertising and on your website. All the speakers had excellent tips about how to increase conversions to purchase. A copywriter advised us to avoid fluffy 'umbrella terms' in the spirit of clarity. A usability expert used heat maps to demonstrate that full-bleed videos on websites were bad, and that many, many online forms from top brands are designed badly.
Web psychologist Nathalie Nahai explored some cognitive hacks from her book that increase conversions, starting with the completion principle. This means we are predisposed to complete things that we start, or have been started for us, which is why it makes sense to autofill forms or give customers loyalty cards with the first few stamps included. André Morys, CEO of German agency Web Arts AG, which is billed as the largest 'conversion agency' in the world, riffed on the behaviour change model we looked at a few columns ago in Admap, which he pithily reduced to 'Reduce friction, raise motivation'.
All of which is well and good and what the audience was looking for. I gave the closing talk and perhaps put some backs up when I suggested that maybe we have overoptimised advertising and marketing to its own detriment. Allow me to explain. The great promise of digital advertising was data. Instead of rough measures of intermediate effect drawn from surveys, we could get persistent, granular, behavioural data of the entire audience, segment them ever more finely, and serve them ads based on what they had done before, rather than what they said.
However, this led to us falling into a quantitative trap. Not everything can be reduced to numbers, which means whenever we consider data, we are susceptible to the McNamara fallacy. Robert McNamara was the US Secretary of Defense, who observed that making decisions based purely on numbers is dangerous. It begins when you measure what can be easily measured, such as all digital things, which is fine. But then we tend to disregard that which can't be easily measured, and then to believe that what can't be measured easily isn't important. Finally, we begin to believe that anything not on our spreadsheet doesn't exist, which leads to over-optimised and underwhelming advertising.
Consider the following tweet, one of the innumerable tweets about how much people really, really hate retargeting: 'Jesus, those f****** Mahabis Slippers are STILL tirelessly stalking me round the internet. I BOUGHT YOU. YOU FELL APART. GIVE UP.' Looking at the data, the conversion marketers would see that after a number of retargeting exposures, there was a sale. This is why the industry continues to retarget: the metrics show it 'works'. But consider the cost of what I call a 'surrender sale'. An angry customer, who now tells all their friends and followers how they had a bad experience with the product, every time they are reminded of it by a continuous stream of advertising. (Yes, there are conversion pixels that would allow the retargeting to stop post purchase, but they are often not used since a previous purchase is a very strong indicator that you are a purchaser, so the targeting algorithm won't give up on you.) What's the cost of that? What metric can represent it?
The problem with endless overoptimisation is that we look at advertising through the lens of a spreadsheet, which inherently reflects short-term growth over long-term frustration with advertising as a whole, which in turn impacts the industry and its long-term efficacy. We lose the humanity in the craft. It leads to ridiculous statements, such as the one recently tweeted by the IAB, the body that defines online advertising. At their recent conference, a speaker felt the need to say: “We need to remember that we're speaking to humans and that advertising is a form of communication.” If there was any clearer reminder of the dangers of the quantitative fallacy for advertising, it's having to be reminded it involves communicating with people.