Artificial intelligence seems like it’s a million miles away, when it comes to having a real impact on the day to day workings of agency life. It’s also easy to be skeptical about AI or machine learning, particularly when everything from toothbrushes or water faucets claim to be “AI-driven” and “AI-infused”.
But is the reality of AI and its impact on the media and advertising industry far closer than we think?
Over the last year, I’ve had the opportunity to see some original research, published by Bant Breen, chairman of Qnary, that takes a hard look at how the advertising and marketing industry is preparing for AI. His November 2019 report, I Think Therefore I Am: An Exploration of Artificial Intelligence and Advertising, surveyed more than 500 ad practitioners regarding their knowledge and attitudes towards AI. How ready are we for AI? The short answer is, not very.
The average self-assessed understanding of AI sits at a very low three out of 10. Many ad practitioners couldn’t name an advertising application of artificial intelligence, and if they did, they couldn’t get beyond Alexa or Watson.
Why is this level of preparedness important? Simple. We’re in a rapidly accelerating era of automation that can affect most tasks in advertising and media.
First, AI is one of those broad terms that act as a catch-all for a variety of machine-assisted learning. True AI and deep learning are rare in any field. Machine learning is far more prevalent and easier to implement. Programmatic advertising and Google search algorithms are both good examples of areas in which algorithms and machine-learning are already embedded in our industry. But the magnitude of the impact of those two examples should be proof alone that we need to think longer-term.
When the research drilled down into individual areas of expertise in advertising and media, there’s a general assumption that anything that’s creative is uniquely human, and anything that is data-driven and repetitive could be potentially replaced by AI. Where did media fall? Nearly 26% of respondents thought that AI could replace media planning, and a whopping 59% thought that AI could play an assisting role. To give you some perspective, media planning is considered to be almost as replaceable by AI as site optimization.
If you as an advertiser are skeptical about AI’s ability to provide higher functioning strategic reasoning and judgement, there is some myth-busting happening on the creative side. Toyota, in collaboration with Saatchi LA and IBM Watson, used AI to create a spot tailored to 100 different demographic targets. The team trained Watson to write the copy based on 50 human scripts – and the end result were insights and language that writers couldn’t have found on their own.
But does that mean AI could automate media planning? Media planning is in many ways ideally suited to AI. It’s based on data, a set of assumptions and rule-based learning, like viewability, wearout, and of course response curves. We’re starting to see new entries into this space. Bionic, for example was launched last year, using AI to provide multiple scenarios quickly and to learn from the human planner’s choices and recommendations.
If a large part of your brain is asking “Why would I train a machine to replace me?”, well, that’s a fair question. But it’s already moving in that direction, so perhaps the better question is, “What does a planner’s job look like when creativity, insight and innovation are the main skills required?”
Short-termism won’t serve us. If we’re to be truly prepared as an industry, we need to manage our way into a world where the craft of media will still be a necessary skill.
Sarah Ivey is founder and CEO of Agents of Necessity