Let’s be honest, Hollywood has warped the way we think when we hear the words artificial intelligence (AI), robot, machine learning (ML), or anything along those lines. Whether they are trying to replace us, destroy us, or even save us, anything in the realm of robotics and AI is not particularly high on the list of things we want to interact with. In our last article, we touched on the future of marketing agencies and the relationship the humans in such agencies will have with technology. In this article, we will explore why we should try to embrace ‘the machine’ and not be afraid of the algorithm.
Firstly, let’s explain what we mean by the algorithm. If we look at the dictionary definition of what it is, it is “a process or set of rules to be followed in calculations or other problem-solving operations, especially by a computer.” Sounds both simple but complicated at the same time, right? Well, let’s break it down further within a marketing context. The people over at Facebook, Google, TikTok, or Snapchat (etc), have made special sets of rules to ensure that you see the content that will interest you based on content and information that you have interacted with before. This includes your activity on their (and other) platforms, your interests, and your habits. For example, if you are interacting with a lot of puppy posts, more puppy accounts will show up on your news feed. Have you ever wondered why people spend hours and hours just scrolling through TikTok or Instagram? It’s all to do with a wonderful man-made creature with a life of its own: the algorithm.
From a marketer’s perspective, the algorithm should be your friend, because whether you are ready or not- the future of this field lies in automation. And yes, a cold shiver also slides down our spines at the thought of being replaced by a cold, faceless entity in Google’s data centers. However, our mindset should be oriented towards thinking of it rather as an enhancement of our human capabilities than a replacement. Before we go any further, we need to know the difference between AI and machine learning as they are often used interchangeably. In the simplest terms, AI is exactly what it stands for: Artificial Intelligence. It is the study of making machines that are inquisitive, learn from their environment, and grow to become more capable of near-human problem-solving. Whereas machine learning is an application with AI that creates a system that can review data and improve itself without being explicitly instructed to do so.
Using AI and machine learning helps to eliminate marketing waste, i.e., it removes inefficient marketing strategies and tactics that are costing your company money. It does this by allowing micro-targeting those consumers most open to your marketing objective in a given moment. It helps you identify the ideal offer, message, call to action in that specific moment. Additionally, due to “the machine” running even when you’re not, you can achieve real-time responsiveness. Targeting decisions and choice of content are done on a micro-level with tens and hundreds of thousands of data points taken into account with each action. Long gone are the days where you’ll have to adjust a bid for a specific demographic group, target location, device, search keyword… the algorithm can do it long before you even have a chance to notice it.
The other great thing about this technology is that it lets you peek into the future through forecasting and predictive analytics. By using ML-powered forecasts, we can make predictions about future behavior which can then influence and guide our marketing strategies on the strategic level.
By having the caliber to pinpoint the exact and most ideal for a specific customer, would give a company a fundamental competitive edge by making sure that they are not spending or doing more than needed to secure ‘the bag’. This inevitably reduces costs as we are eliminating waste that occurs naturally due to mass marketing by being more mindful (with the help of our algorithm buddies) and customizing each offer with the minimum amount of influence to convert purchasing decisions. With ML, we are also able to streamline marketing efforts, as we can predict which segments of the population are most likely to become customers. This is because ML goes beyond traditional data analytics as it naturally integrates (and builds upon) the results of previous marketing efforts and uses them to improve future decision-making. Moreover, echoing what we already said, using AI and ML lets us do advanced personalization. The Boston Consulting Group measured a 10% increase in revenue in companies that have embraced personalization tactics in their marketing efforts. They also predict that over the next few years, there will be an even starker difference between companies whose personalization ranges from little to none, and those who do it just right.
Now that we have established why AI, ML, and co. are our friends, let’s look at what we need to keep in mind when working with them. The first point concerns trying to manually intervene in particular points where optimization algorithms already have plenty of data on. If you are doing this, as a rule, we hate to break it to you, but not only is it a waste of time, it’s also pointless. This is because the algorithm will look at hundreds and thousands of more granular user-level data points, and do it regularly. HOWEVER, there is an exception of when you should butt in, and that is when things like ethics or product specificities and outliers come up. This is because ML does not understand them very well (yet). If you would like to avoid any faux-pas in the future, you need to make sure to intervene at these points to make sure that your ML orientates your marketing efforts that align with your brand and its image.
Secondly, the algorithm is unfortunately not a well-kept secret, and it’s available to everyone in the biz. However, this does not mean that you should just give in; as we mentioned in the previous article, we need to feed the algorithm only the creme-de-la-creme of information so we can maximize the benefits. What this means is that you should give your algorithm with the ‘right’ data, based on what you already know about your business/industry/users/conversion rates/objectives, and so on. Lastly, the algorithm, like us, is not perfect, and so it will not always be 100% correct, as it can only base decisions on previous data. If you feed an algorithm with irrelevant or not enough data, you can’t expect good decision-making; in computer science, they refer to this as the “garbage in, garbage out” principle.
And so to conclude, whilst the algorithm is your friend, it looks like that it will take over the marketing monarchy, be ‘King’, and potentially become ‘Emperor’ as its capabilities improve. Some of the points in the book “The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World”, written, by Pedro Domingos and published in 2015, could have easily been scoffed at. The truth of the matter is that each month and year since then, we have more or less quietly witnessed huge capability improvements of the major AI/ML-based technologies. As marketers, all we can do is make sure that it is fed a healthy diet of correct data, intervene only on necessary points, and just hope that our efforts are not the catalyst for a Hollywood-esque robot uprising.
Ivan Lucic is one of our marketing specialists at Scalista. While he spends most of his working hours invested in search campaigns, the intricacies of app & display certainly pique his interest as well. Outside of work hours, Ivan enjoys gaming, good food, and exploring the world.
Megan Joye McFadden
Megan is our marketing assistant at Scalista. She is biotech student who (kind of) switched lanes over into business, and has a flair for all things creative. When she is not assembling all the expertise from our team into a coherent blog, she can be found dabbling in all kinds of activities that involve podcasts, science communication and baking. If you would like to get in touch with our black-tea-with-milk-and-sugar-drinking blog writer, you can find her on LinkedIn.