Manual over-ride must be an automatic option

Marketers must get under the skin of automated decision-making systems so they can help them make the right choice.

Marketing is a heroic profession. Or so we like to think. Compared with our colleagues in, say, finance, IT or manufacturing, we have an image of ourselves as courageous change-makers.

Ever since Theodore Levitt, perhaps even before, we have been cast as creative visionaries who possess prophetic insight into our customers; the whiz kids whose role it is to organise the rest of the business to meet the opportunities of the market.

However, the reality is that we are often little more than well-paid machine minders. The sophistication of one-to-one marketing technology means that much of what we do these days involves managing complex, semi-autonomous decision-making systems and contact strategies.

And surrendering authority to a machine can make us nervous.

A few years ago I took a call late at night. A colleague with whom I was due to run a workshop the next day was running late. He’d refused to board a plane in Los Angeles. The plane in question was an Airbus A340 and several weeks before he had watched a Discovery Channel documentary in which a similar aircraft had ploughed into a forest at the end of a runway.

My colleague is a rational man. He believed that the Airbus family of airliners is intrinsically less safe than those made by Boeing, for the simple reason that the flight controls are automated. A computer sits between the pilot’s hand and the control surface, interpreting the human inputs and second-guessing any instruction that the many microprocessors believe to be dangerous.

Many people in the aviation world agreed with him. Some still do. There is something unsettling about being second-guessed by the very machine one is trying to control. De-humanising, even.

I am just about old enough to remember perhaps the best piece of flying ever. In July 1969, Neil Armstrong wrested control away from the computer managing his descent to the moon. It was taking him towards a boulder field which might have destroyed his craft. He landed with just seconds of fuel left in the tanks.

We’d all like to be Neil Armstrong, winning one for humanity in the struggle against the machine and vindicating the importance of “the man in the loop”. It’s this reptilian bravado you hear in the voice of the motoring journalist, decrying the nanny manufacturer who presumes to make their stability management system so hard to turn off.

The hedge fund manager didn’t come across as a seat-of-the pants flyboy. He is a rational man with a great respect for the data that complex information systems provide. Indeed, he claims it was his peers in the banks who were the flyboys. They willfully ignored what the data was telling them about the banking crisis on 2008.

In marketing, the equivalent is the heroic bank marketing director who rewrites every customer communication in the pursuit of his new brand strategy. And in so doing destroys the benefit of two decades of direct response testing and learning.

The other week, my boss and I sat down to dinner with a hedge fund manager, who had predicted the banking crisis of 2008. (In case you’re wondering, he paid for the meal.)

Our host didn’t come across as a seat-of-the pants flyboy. He is a rational man with a great respect for the data that complex information systems provide. Indeed, he claims it was his peers in the banks who were the flyboys. They willfully ignored what the data was telling them. He saw the truth in it.

He was enthusing about a book that described how someone had used an approach very similar to his in an even more challenging set of circumstances. Fly By Wire by William Langewiesche is one of the best books about flying since The Right Stuff. And perhaps the best ever book about humans and machines.

It tells the story of how, in January last year, Captain Chesley Sullenberger saved his aircraft, passengers and crew by landing in the Hudson following a birdstrike. It is a story of heroism, but not in the analogue Neil Armstrong mode. It is a story of humans working together with automated systems to save an aircraft. And yes, that aircraft was an Airbus.

“Sully” Sullenberger did many things right that day, but three actions stand out. First, he went into the crisis knowing he would need all the help he could get from the aircraft computers. So as the engines went off-line, he powered up the auxiliary power unit to ensure an uninterrupted electrical supply.

Second, Sullenberger let the flight control system hold the plane straight and level while he took the time to think. Rather than asserting immediate control, Sully let the computers do their thing while he and his co-pilot focused on the kind of problem solving that only the humans in the cockpit were qualified to do.

Last, Sullenberger used that thinking time to make the kind of creative decision no computer would ever make. Rather than diverting to either of the two available airports, he elected to put the plane down in the river. A course of action which analysis later showed was the safest of all the available options, if not the most conventional.

It seems to me that in the actions of our hedge fund manager and Captain Sullenberger, we see the emergence of a new kind of decision-making model for the information age. A blended approach in which the best practitioners get under the skin of the systems and work out how they can help them perform better as human decision-makers.

This may sound a little less than heroic, but as bush pilots in the far north of Canada once said: “There are bold pilots. And there are old pilots. But there are no old, bold pilots.” The same may one day apply to marketers.

Richard Madden is planning director at Kitcatt Nohr Alexander Shaw


    Leave a comment