Several years ago, Brian Solis (@briansolis), a Global Innovation Evangelist at Salesforce, observed, “There is no longer a delineation between digital and traditional customers. Anyone with a smartphone traverses between online and offline activities without a second thought. As they do, they gain a penchant for modern conveniences, such as speed, utility and real-time assistance. Along the way, they also become more and more impatient and demanding. … Leading marketers are using the likes of machine learning and other emerging technologies to deliver assistive experiences that convert intent and expectations to value-added engagement at every step of [the consumer’s] purchase journey. “[1] Thanks to the pandemic, the number of consumers taking the digital path to purchase has increased at a faster pace than had been predicted when Solis wrote those words.
About the same time Solis was making his observations, Neil Lokare (@NeilLokare), currently a full stack software engineer at Opendoor, observed that many marketers are concerned they aren’t ready to use technologies like machine learning. He explained, “Cutting-edge advertising technology is often marketed as ‘powered by Artificial Intelligence.’ Some marketers get scared away, believing that they aren’t prepared or educated enough to leverage the latest tech. However, many marketing and advertising applications of Artificial Intelligence (AI) lean heavily on a subset of AI called Machine Learning. Essentially, Machine Learning (ML) is about improving predictions from the data you already own. As a decisionmaker within your organization, you should understand what basic ML applications entail and how they can be properly employed.”[2] He believes with just a little background education, marketers can take full advantage of machine learning.
Evangeline Sutton (@Evangel_Sutton), a Marketing and Sales Executive at Regenerative Marketing LLC, agrees with Lokare. Rather than being apprehensive about machine learning, she believes marketers should be excited to learn more about it. She explains, “This is an exciting time to be in marketing. Instead of losing ground, you can stay ahead of the curve by staying up with the changes and being willing to implement the additional tools. Create a platform that feeds you news and set aside a small amount of time to educate yourself on how to use it for your clients each week.”[3]
Machine Learning in Marketing: The Basics
The staff at Google prepared the following video that lays out three steps every organization must take in order to get the most out of machine learning. Those steps are: Collect, analyze, and activate.
During the collect stage of machine learning, Gyi Tsakalakis (@gyitsakalakis), founder and president of AttorneySync, cautions, “Be careful about the data you input. Not to throw cold water on the party, but I encourage advertisers to approach new technologies — particularly machine learning — with a healthy skepticism. While there’s little doubt that machine learning will play an ever-increasing role in what we do, it’s often oversold as a panacea. If you train your machine on garbage data, it will learn to deliver garbage.”[4] In other words, make sure you are collecting the right data and using it the right way.
During the analyze stage, the video notes, “You’ll want to build an accurate model you can trust. Start with an outcome you already know to test for accuracy.” But what kind of model? Lokare discusses several types of models: Regression, classification, and clustering. He writes, “Regression is the process of estimating the relationship between variables. For example, a basic Return on Ad Spend (ROAS) model: given $100 of ad spend (X), how much revenue can I expect to generate (Y)? … Predicting ROAS could be a function of budget, time of day, average frequency caps per user, site content ratings, ad visibility, ad size, button color, ad copy, etc. This is known as multiple regression.”
Concerning classification models, he writes, “Classification is a form of categorization via pattern recognition. At its simplest form, logistic regression is the model of choice for determining the likelihood that something is either Category A or Category B. … A more relevant example for advertising would be a classifier that determines the probability of a user converting after seeing an ad. This type of classifier powers the automated bid optimizers found in many [demand-side platforms (DSPs)]. … The higher the likelihood of conversion, the more your DSP is willing to pay for the impression and thus a higher bid. However, not all classification has to happen within an auction.”
Finally, concerning clustering, he writes, “Classification and regression are typically supervised learning models, meaning the input data is labeled for a particular set of outputs. Clustering algorithms are unsupervised — the process of finding patterns and hidden structure within unlabeled data. Think of unsupervised models this way: ‘What patterns can I find in a set of data that I don’t already know much about?’ Audience or customer segmentation are prime examples of applied clustering. Your Customer Relationship Management (CRM) system is filled with a breadth of customer-related information. Wouldn’t it be valuable to figure out who you should reach out to for a customer loyalty promotion? Or, what kind of client is likely to churn and what can you do to retain their business? Clustering algorithms can identify the similarities between your users/customers and group them accordingly.”
During the activate stage, the Google video concentrates on customer segmentation. It says, “This is where you’ll take the output and use it to segment your customer base into high, medium, and low risk tiers, providing each with a customized retention offer. This will help you determine where and how to spend.” Clearly, you want machine learning insights to help you determine where and how to spend; however, it can help you do more than simply segment customers. Oz Etzioni (@ozetzioni), CEO and co-founder of Clinch, “When it comes to marketing, we can divide AI to three main aspects that can help us better measure and understand how, when and to what extent to utilize AI and the value it can provide: Automation, Decisioning, Optimization.”[5]
Concerning automation, he writes, “AI allows us to streamline the production of ad experiences and messaging at high scale, quickly and cost-efficiently, using dynamic creative templates and massive feeds that include all the potential combinations of products, messages, locations, call-to-actions, backgrounds, content, and more, where literally every component in the ad experience is fully dynamic and can be customized and personalized per viewer in real time.” He continues, “‘Decisioning’ is where data is introduced into the mix. How do you make sure that every element and component within the experience you serve your consumers is the most relevant for them and resonates with them? How do you make sure it relates to their own personal context, location and time — three important factors that can build brand trust or crush it instantly if not done and executed correctly. AI plays a huge role here in the real time decision making of what to show to who, when and why.”
Finally, concerning optimization, he writes, “‘Optimization’ is where scale and speed matter the most. Just think, not that long ago the first machine-based optimization was just simple A/B testing for better click rates, then media performance was added to click rates and that made for more effective spending, but you were still measuring only two variables … but now, there are dozens of variables and more complex decisions, can we handle it the same way, at the same speed we used to for a single or two variants? Well, as a marketer you already know the answer.”
Concluding Thoughts
The experts point out that machine learning in not a silver bullet or a panacea — it’s a tool. Used correctly, it’s a powerful tool. Ben Jones (@harperjones), founder of Google’s Unskippable Labs, explains, “Using machine learning is like having a billion interns working for you, not a single Einstein coming up with the perfect solution. You have to figure out how to use them, which requires assigning them tasks and translating their output into something useful. Without you, the interns would be lost. Rather than belabor the industry debate of creativity versus technology, I think we creatives are better off figuring out how to guide the interns. When we harness machine learning’s ability, we’ll make better, more relevant, and more effective ads.”[6]
Footnotes
[1] Brian Solis, “Marketing In The Age Of Machine Learning,” Forbes, 8 November 2018.
[2] Neil Lokare, “Machine Learning Basics that Marketers Should Know,” MightyHive, 13 February 2019.
[3] Forbes Agency Council, “10 Tips For Leveraging Machine Learning In Advertising,” Forbes, 21 December 2018.
[4] Ibid.
[5] Oz Etzioni, “Artificial intelligence (AI) in marketing,” ClickZ, 14 July 2020.
[6] Ben Jones, “Machine learning is like having a billion extra interns, not one Einstein, to make effective ads,” Think with Google, June 2019.
