Fear of Change

There’s been lot’s of chatter (and quite frankly panic) recently around the supposed danger AI poses in the marketing effectiveness industry. Like any tool, improper use can lead to less than desirable outcomes: but that shouldn’t stop change from happening. If used properly, AI can deliver a whole host of benefits and overcome many of the flaws that currently exist within our industry.

It’s a well established fact that throughout history industry is generally slow to accept technological change. Whilst seeking investment in his then fledgling motor company in 1903, Henry Ford’s lawyer, Horace Rackham, found that he encountered stiff opposition from those that believed that the tried and trusted use of the horse and cart worked perfectly well (with the president of the Michigan Savings Bank infamously stating “The horse is here to stay, but the automobile is only a novelty – a fad”).

More recently in 2007, on the launch of the very first iPhone, Steve Jobs faced similar criticism from many of the established industry players who’d simply failed to understand just how revolutionary the device would be (with the then Microsoft CEO, Steve Ballmer famously stating “There’s no chance that the iPhone is going to get any significant market share, no chance”).

Enhancing Effectiveness and Accessibility

The advent of AI and it’s use in the marketing effectiveness as well as other industries has once again elicited a similar response.  Culminating in a rather hysterical open letter to Marketing Week (Marketing evaluation experts warn of dangers of AI (marketingweek.com), many of the current self-styled industry “heavyweights” seek to undermine what could be a potentially revolutionary advancement in the field of marketing effectiveness. Born out of a potential fear of change and a possible desire to protect existing business models, the article overstates the potential flaws of AI whilst failing to mention any of possible benefits.

Implemented correctly, AI has the potential to greatly enhance the depth and speed at which marketing effectiveness analysis can be conducted. This process “efficiency gain” will ultimately mean that marketers can gain access to richer, more granular insight when they need it, as opposed to the current situation whereby analysis is often too high level and provided too late in the marketing process to truly be of use in marketing planning.

The letters authors do, however, make a valid point in that the use of AI should never be “black box” and always contain a human component. Whilst I agree in principle with this statement, the authors fail to point out that this criticism can equally be levelled at existing effectiveness providers, who have historically been very reluctant to reveal the process, assumptions and underlying models used to conduct their analysis.

Given the apparent consequence of using so called “bad models” in decision making (according to the article, up to 40% of media driven revenue is supposedly lost via a “bad” model), all providers should be required to prove the quality of their work by being transparent about the approach taken and assumptions used, working closely with both the client and independent experts within the industry they are developing the models for to deliver the best outcome.

Indeed, AI affords an opportunity to again improve on established practice here via the ability to weave in the “rules of the game” when building a model. By teaching the algorithm to build models that align with more intuitive tests around model structure and result benchmarks, the quality of output can be improved to the extent that it eclipses those produced by many of the more established providers, greatly enhancing standards within the industry.

Finally, and perhaps most significantly, AI presents the opportunity to provide greater access to marketing effectiveness amongst those clients for whom advanced analytics was previously thought to be out of reach.

By improving efficiency the data requirements and subsequent cost of entry are lowered meaning that a greater range of clients in a variety or marketing disciplines can gain access to data led decision making. This has the potential to vastly improve the overall quality of marketing decision making and execution, generating greater returns for the clients business which, ultimately is the point of conducting such analysis in the first place.

Given these obvious, tangible benefits, the effectiveness industry should therefore embrace the advent of AI as opposed to fearing it: After all, when faced with a choice between an iPhone or a Blackberry or indeed an automobile and a horse and cart, if I were a client, I know which one I’d choose.