Marketers who look for artificial intelligence to provide fresh insights don’t always get what they want. Insights are not always easy to categorize, let alone evaluate and implement. Often patterns emerge which may be intriguing but unprofitable to pursue.
A fast-growing collection of ambiguities, gray areas, and information that demands deeper levels of analysis are likely what the face value searches give. Since marketers began to use AI several unintended consequences of significance have emerged. Marketers have had to make decisions about ethics.
AI can play a huge role in classifying humans based on assigning labels on what AI sees in photos and videos, such as skin color, hair, religion and more, which adds bias in marketing. With AI, marketers have had to decide between brand and performance.
When using AI, many times insights are outside the human-created brand guidelines, personas or practices. Now, marketers are facing the challenge of ignoring these AI insights or testing the data and using those insights to churn results.
AI is forcing marketers to redefine themselves. In every marketing team, there’s a data and analytics team, which usually makes up a small group of analysts that are crunching the numbers.
Marketers pride themselves on driving strategy and being creative, with analysts backing up those decisions with data. But AI provides a new abstracted layer from strategy to performance, forcing marketers to redefine their relationship with data.
Some marketers rely on data as a ‘fact check’ for gut-based decisions – refreshing a dashboard and doing a cursory glance at the analysis to validate their preconceived conclusions. The unintended consequence is a vicious cycle, promoting more of the same.
Instead, marketers must do more than “check the box” on data but instead become data experts themselves – digging in, challenging assumptions, asking strategic and creative decisions of the data.
The world has evolved to a point where AI can be used to shape marketing strategy, not just creative or campaign execution. How should this capability be assessed and managed? First, assess the inputs. A machine is only as good as its data.
Start with assessing the information going into the model. Marketers should do a data audit to confirm consistent asset tagging, and validate syntax, valid metrics, data formats before starting. Understand how often the data will be refreshed to ensure the AI is working on valid assumptions and timeframes. Create an AI ethics statement.
An AI ethics statement should be a living, breathing statement that is evaluated and iterated quarterly as our experience and real-world examples shape our understanding of best practices. It should cover three areas, data collection methods, tenets of the statement (methods by which ethical use are being evaluated) and data use.
In the marketing world, what is AI not doing that it should be doing?
AI should be focusing not on the “what” but on the “so what?” Much of artificial intelligence in marketing is focused on delving into massive amounts of data, finding unique patterns and outliers invisible to the human eye, and then delivering that specific insight.
So what? Artificial intelligence should be taking marketers into the steps of past analysis, into recommended options and their predicted outcome. For example, AI finds a new, problematic anomaly in a massive data set – what are three options for potential resolution, and the calculated outcomes associated with each?
With both the new outlier and its anticipated consequences calculated, AI has done the double work of calculating both problem and solution – and the human can be best used in applying judgment, ethics, and creativity to solution deployment.
How are AI tools and the human touch best blended?
According to R. J. Talyor, Founder and CEO of Pattern89, an AI firm that helps marketers predict their campaigns’ performance, “Humans provide new thoughts, creativity and imagination. Artificial intelligence can identify patterns and outliers invisible to the most skilled statistician.”
A human marketer is essential as the creative input and the brand and ethical judge of the output from the AI. It’s a perfect pairing marketer & machine. “AI will reinvent marketing, again. Just as the internet (1993) and the iPhone (2007) forced marketers to fully reinvent their strategy to adopt digital and mobile, respectively, AI will revitalize creativity and brand differentiation,” he noted