Facebook has started running advertisements recently in conventional, analogue media. Amazon is busily buying and opening physical shops. Perhaps it's time the advertising industry awoke to a similar counter-revolution.
This is what I suggested in a keynote speech at Unbound London recently. As a 53-year-old Conservative fat man, I always enjoy giving slightly controversial talks, since the maniacally politically correct have turned platform speaking into a high-adrenaline extreme sport, like cave-diving or BASE-jumping: from the moment you clip on your lapel-microphone, you have the thrill of knowing you are only one misjudged pronoun away from career suicide.
But, even more than that, I enjoy making mischief with the technology industry. Not because I have any animosity against it, but simply because the tech industry has long seen itself as the bloody answer to everything. As a result, it is fun to remind such people that in every problem field there is a line where technology ends and human psychology begins.
In my talk I referred to what I call "the Valley Vice". At its worst, I explained, what Silicon Valley loves to do is to automate those parts of a system most susceptible to automation while ignoring the wider whole. More broadly, this approach rests on the fallacy of assuming you can optimise a complex system by separately optimising one of its individual parts.
For example, a technologist might see a hotel doorman, define his function as "opening the door", replace him with an infrared automatic door-opening mechanism and declare victory for superior efficiency. The only problem with this approach is that a hotel doorman has many other roles besides opening the door. He helps with luggage, hails taxis, affords recognition to guests, and lends a measure of status and security to the hotel. None of these functions is performed by an automatic door.
Six months later, what your automatic door has saved in salary costs may be more than outweighed by the fact that the hotel's reputation is in freefall and guests have abandoned it – perhaps because there’s now a wino asleep in the doorway.
Again, the mistake is to pretend you can take a component of a complex system, define its function very narrowly, automate that function and then reintroduce the automated component to the system, confident that an improvement in one part will lead to an improvement in the whole. It simply ain’t so.
Yet this is exactly the trick that technology companies – with the connivance of media agencies, I might add – have successfully pulled on the advertising industry.
By deluding everyone that the whole of advertising is reducible to "the efficient and inexpensive delivery of targeted messages" through the extensive use of data and algorithms, two companies have gained a multi-billion-dollar rent-seeking monopoly over the majority of advertising activity.
The simple fact is that Facebook and Google cannot realistically claim to have a monopoly on creative ideas. So they grotesquely overstate the relative importance of those parts of the business where they do enjoy a natural monopoly. Clients, keen to demonstrate quantifiable incremental improvements, willingly go along with this fiction. Media agencies, understandably eager to raise the status and complexity of what they do, connive in the deception to form a kind of Bootleggers and Baptists coalition with digital media owners.
Do you think creative agencies are even involved in creating most of the ads you see online? Are we fuck. The media agencies produce it themselves as an afterthought.
But this whole targeting-obsessed scientism is far from being the only game in town. I would not deny for a moment that targeting is vital to many areas of marketing communication. But it is nonsense to treat advertising with the mindset of a kind of digital Taylorist, and to ignore the wider context in which messages are processed and in which brand trust is established. Like a hotel doorman, advertising performs many jobs, only a few of which can be automated.
Most marketing problems are what I like to call Sudoku problems. What this means is that any solution can only be obtained by considering the system in its entirety. You can’t cut a Sudoku into nine separate squares and ask nine separate people to solve one square each.
And, in the same way, you won’t reliably maximise the long-term profitability of a business or a brand by encouraging different groups of people to focus exclusively on improving one thing to the exclusion of anything else. That’s why it never made any sense to me that media planning and creative should be separated at all. Instead of going to clients saying "look, we have a solution to your Sudoku", we are now reduced to saying "we think the number in the top left corner might be a three".
Nerding around with data and media efficiency is a sideshow – merely one corner of the puzzle. As every survey of advertising has found, the bulk of the value created by the marketing services industry is far from being a question of targeting alone but requires distinctive creative ideas in the service of strong brands, widely disseminated, often in media perceived to be expensive.
We should surely be proud of this creative alchemy rather than shamefacedly pretending we are becoming a branch of the data science industry. The fact that management consultants want to buy us terrifies me. I mean, have we really become that boring?
Rory Sutherland is vice-chairman of Ogilvy