Advertising effectiveness: A complex problem

Anyone who knows me may suspect at this point that this is going to be a pedantic semantic distinction of language, since I enjoy those. But it’s actually a good deal more important than that, because it goes to the heart of what we do every day.

And we’ve been getting it wrong as an industry.

A complicated problem involves a range of variables, steps and processes to get to a good outcome, but each variables will respond in predictable ways. So if you create a good system, apply checklists to it and repeat, you will generally have success in managing that problem. Software development is a good example of this. Loads of lines of code that need to work together to produce an outcome. But if you’re skilled and rigorous, the software will do what you need. It’s complicated, but there’s a right answer.

A complex problem is much less predictable. It contains a lot of uncertainty. And the variables are not all under your control. Trying to apply a process to it can create the illusion of certainty, but it’s never anything other than an illusion and can often lead to counterproductive results, making the problem even worse. The way to manage these problems is to embrace the uncertainty and focus on creating good outcomes, rather than over analysing all the components of that outcome and trying to manage them individually as though there were a "right" answer. In complex problems, there are only outcomes, not answers.

Which brings me to the topic of advertising effectiveness.

It’s a complex problem. But everyone wants to manage it as though it were a complicated problem.

There is a broad mentality that says that if marketing and advertising people were any good, and not just the colouring-in department, they would have been able to quantify how to make advertising work predictably by now. Not doing so is simply flaky people trying to write themselves a blank cheque for spending loads of money on self-indulgent creativity. Try harder!

And, of course, as agencies we will often fall into this way of thinking in an attempt to curry favour with the clients that matter (the ones who decide where the money goes). We build econometric models and attempt to quantify creativity with pretesting systems. But we’re just trying to manage a complex problem as though it were simply complicated.

Tech companies understand this.

Whether explicitly or implicitly, they realised early on that advertising effectiveness would be easier to manage if they reduced the variables. So, until very recently, they turned effectiveness into a simple problem. Virtually all the metrics they ever reported were ones in their control. Serve someone an ad, ask that person how they feel about the brand while they’re on the platform. Serve someone an ad, work out how long they spend with it on the platform and report that. Serve someone an ad, measure how many of them click through to your website to buy. It makes advertising with them feel robust, accountable and predictable. It solves the question of advertising effectiveness if the only advertising you’re doing is with them and the only outcome you’re interested in is what happens on their platform.

But we all know that, for most big brands, effectiveness is more complex than that, because they’re using more than just Facebook. Anyone who has ever been involved in trying to create an econometric model knows the wide range of potential variables that are entirely out of our control that affect the performance of our advertising. Even econometric models can’t solve this complex problem; by creating the right answer, all they can do is reveal the underlying effect of advertising that is otherwise hidden under the complexity of the real world.

We can thrive in this environment if we stop apologising for complexity and instead educate others in our businesses that just because something is complex doesn’t mean that it can’t be effectively managed or produce great outcomes.

Not everything that is valuable to a business is predictable. Predictability is not the state to which every department of a business should aspire.

We should embrace complexity and adopt a complexity mindset to manage it: invent, try new things, learn from them, do it again. Worry about whether the outcome is good rather than obsessing about the precise efficiency of each of the inputs.

Advertising effectiveness is about orchestrating a set of complex and unpredictable elements to produce a positive outcome for the business. This outcome can be dramatic and hugely valuable. If we instead treat it as a complicated problem, we will measure it better by neutering its effectiveness. Let’s not do that.

Craig Mawdsley is joint chief strategy officer at Abbott Mead Vickers BBDO.