Entry 1. February 18, 2021
By now, we have all heard that AI is revolutionizing marketing and sales. Whether automating traditionally manual processes, using marketing analytics to predict future behavior, or even using it to help select what message resonates with which of your customers, AI is gaining more momentum every day. The question is, are you – as a marketer – ready to embark on the journey?
It may seem daunting, but like anything else, you have to choose your focus and master one thing at a time. While I am the COO (and veteran marketer of 20+ years) of an AI design and development firm, I assure you that I came into the world of AI and data science with way more questions than answers. Over the past 4 years I have learned enough to be dangerous and developed a knack for translating “techie” into language that resonates with our audience of innovative business professionals; especially the marketing message. So when we saw a new role emerging in the industry, it really spoke to me. That’s when my journey to become an AI Translator really began.
I want to write this journal series to chronicle my “quest for AI enlightenment” in hopes that it will resonate with the challenges that many of my fellow marketers are facing. There are so many tools to do so many things, there just isn’t enough time or budget for most of us to use them all; so we need to choose wisely.
But before we start looking at marketing AI and its associated applications and tools, I want to talk about creating a foundation for AI success…and that starts with ethics. Let’s face it, marketing and sales are two of the most human-centered applications of AI, and we as marketers need to be stewards of fairness.
Data doesn’t lie, but it can be misinterpreted, and that is why we hear so much about bias in AI. When considering the implications an algorithm can have on a wide swath of people, we need to understand the purpose the data you are using was originally collected for. Does it align with the audience and intent of your marketing initiative? Training an algorithm on solely your own data can also lead to bias, as it is likely not representative of the entire market. For example, training a lead qualifying model on a predominantly male customer base can miss out on the female demographic, or even worse, create models that alienate them. As marketers, we all know the pitfalls of appearing skewed and the implications for the reputation of our brand.
While I am just scratching the surface of marketing AI in this first journal entry, I will stop here for now as I want these to be easily digestible. My next entry will build upon the notion of ethical AI, discussing mitigating bias to reinforce diversity and inclusion. I hope you’ll join in on the journey and learn along with me!
Nicole Ponstingle is COO & AI Translator at Pandata.