All you need to know about AI marketing
Whether we realise it or not Artificial Intelligence (AI) is now a major part of our daily experience. Internet giants such as Google, Facebook and Amazon are all AI platforms guiding what we see and when we see it, thus having a huge effect on our decision processes, especially when we make purchasing decisions.
For business owners and marketers, the question is am I going to passively allow AI to be a powerful tool that my competitors take advantage of while I steer clear because I don’t understand it, or am I instead going to find ways to harness, control and benefit from it?
What is AI?
A traditional definition of artificial intelligence, as defined by Alan Turing and put to the test through The Imitation Game, is the ability of machines to replicate human thinking and reasoning. In reality though, AI in industry is actually quite narrow, much more about replacing monotonous tasks with precision but we also see more advanced applications in practice in our everyday lives such as when our navigation apps help us to avoid a traffic accident or Google helps us find information on anything that we happen to be looking for, or when Gmail finishes our sentences and also when Siri responds to our questions. Even the more advanced applications are still what is classed as narrow in comparison to AGI or artificial general intelligence, which is when machines become capable of and even surpass human level intelligence.
In most applications the underlying AI model or algorithm is processing huge amounts of data or information using a descriptive analytical function and then processing it’s learning through to using a predictive analytical function to predict an outcome and direct us to the best possible solution.
This speaks to one of the fundamental principles of marketing, namely identifying what it is that our customers need or want and how our product or service can satisfy these needs or deliver solutions to their problems.
Data is of course essential to this whole process, however, raw data on its own is of very little use to anyone whether human or machine. In order to be of use it needs to be linked with informative tags, or in AI terms ‘labelled.’ At its most basic level this can be demonstrated by having processed billions of image searches for the term Jaguar, Google is most likely to show you images of the car company’s most popular models as opposed to showing you images of the jungle cat as it now knows that for all those people who searched for that term the larger percent clicked on a particular image or source showing the car rather than the cat. Google can understand sentiment, which in turn leads to correct predictive outcomes.
This has extended to purchased versus didn’t purchase, clicked versus didn’t click and, following the purchase of YouTube in 2006, watched until the end versus switched videos. For Google, it has been reported that AI has always been the intention. When ‘digital prophet’ Kevin Kelly sat down with the founders of Google in 2002 to discuss their free to use web search business and questioned how they expected to make money, their response was simple: “Oh, we’re actually building an artificial intelligence company.” They knew the vast amounts of data that would flow through their platform would ultimately lead their creation down that path.
How to make use of it?
Having worked in this field since 2005 and having seen how systems and platforms have developed, I can say that it’s now time that we stopped considering these internet behemoths as simply advertising platforms and more as AI super engines. To draw a direct parallel from another area of AI let’s consider air flight. Every new commercial aeroplane in the world today is driven and controlled by a computer for 98% of the time, with the pilot there to make the important and critical movements at the right and appropriate time. Every business now needs an AI pilot there to guide the trajectory and journey of the marketing campaigns and make the critical decisions when they need to be made, all the time working in harmony with the AI to deliver better results.
A large part of this requires taking a leap of faith and relinquishing human control. Whereas in the past undertaking a major advertising campaign would require the assistance of a large media buying team or agency that would assess different platforms and positions and recommend which ‘spots’ you should purchase, often making a hefty mark up in the process. AI engines such as the Google Display Network (GDN), if set up in the right way, will do all that work for you. As described in the opening sequences of the 2019 Netflix film, The Great Hack, it now appears that our devices are listening into our conversations as more and more of us experience advertising being displayed to us at points in time that seem to be very coincidental.
A more precise explanation is that our behaviour is being accurately predicted based on our likes, shares, interests, search history, and thousands of other data points that the platforms collect when performing its descriptive analytical functions before moving to performing its predictive analytical functions. Here, the GDN not only takes on the role of media buyer, matching the ad to the context of the web page, but also identifies the interest of the individual reader visiting the page. The magic of AI then kicks in with tailored, personalised ads appearing the moment the visitor lands there. Based on the enormous amount of data points attributed to each and every one of us, a platform such as Google can predict which ads are of distinct interest to us, and they do this with uncanny accuracy. It does, however, get rid of the reassuring knowledge that your advert will appear in a given location, at a given time. With the majority of AI driven campaigns generating very healthy returns however, this soon becomes much less of a concern.
What is the future of AI?
To put it simply, the future is now. The number of data points and AI applications are advancing at a ferocious speed. Acclaimed author and former head of Google China, Kai-Fu Lee, identified four waves of the AI revolution in his 2018 book AI Super-Powers. The first two, Internet AI and Business AI have already been addressed extensively in this article. The third wave, Perception AI, we’ve touched upon and is the digitising of our physical world. Smartphones now use Perception AI to recognise our faces and protect our devices and wallets, while Amazon Echo is digitising the audio environment of our homes. This is set to become ubiquitous, not just limited to certain devices and platforms but all encompassing, providing even more data and touchpoints to the great AI engines. Virtually everything is set to be connected, with people moving between a physical and digital world simultaneously hundreds of times every day.
The final wave is Autonomous AI and is made possible by machines being able to see, hear and react to the world around them. We can see this already gaining ground, although very much in its infancy, with the development of self-driving cars and hazard awareness and prevention. Another way this will be harnessed is with fully autonomous drones performing a multitude of tasks. Just imagine a “swarm” of intelligent autonomous drones working together to extinguish forest fires. In today’s world, we can surely see how this adoption will be of incredible value when it becomes possible.
As this journey unfolds, machines will have the ability to optimise their performance from being immersed in our daily lives leading to enhanced learning from extremely complex data sets that expands continuously, “fuelling” the algorithms allowing them to understand and interact with the world more so than ever before.
AI is no longer the work of science fiction and prediction; it's a very real part of our lived experience.
While most businesses are not in the position of developing self-driving cars and autonomous drones like Google and Amazon, they still need to take advantage of those AI applications available today that will help to better market and grow their businesses. Over the next few years AI will decipher an unimaginable amount of information collected by billions of sensors embedded into our cars, draped over our bodies in wearable smart clothes, and in our homes and public spaces. Those businesses with the right expertise in place are set to benefit from unprecedented targeting and personalisation, but to do so need to embrace AI now.
