From luminaries like Stephen Hawking and Elon Musk to lesser known skeptics sporting tinfoil hats, artificial intelligence has inspired much concern in many. It’s exciting, disruptive and, as these skeptics are at pains to point out, potentially dangerous. A lot of talk has been made about how the robots are coming to take our jobs. As industrialization threw countless artisans and tradespeople into unemployment, so to will AI make many white-collar professional obsolete. Or so the line goes.
At Coherent Path, we have a bit of a different perspective. From our own experience delivering machine learning marketing applications to retailers, we’ve found that AI and automation actually help marketing and creative professionals do their job more effectively and even provide new opportunities for work. Here’s how…
1. It helps identify the right subjects to focus on
Before email campaigns, ads and website product descriptions can go out, the requisite creative needs to be prepared: photography, copy, graphic design and, increasingly, video. But which products and topics should you focus on?
Usually, the marketing department decides what to feature and emphasize based on questions such as:
- What does merchandising need?
- What attracts people most about this particular product?
- Who does this resonate with?
- How did similar products or campaigns do in the past?
- What is appropriate for the season?
Once the marketers have answered these questions and weighed the priority of each, the graphic designers, photographers and copywriters are called in to put together the creative that will entice customers into clicking and converting.
However, it can be exceedingly difficult for marketers to determine who will like what and get it right every time. Inevitably, some products won’t resonate with certain people–no matter how well put together the creative selling those products is. Marketers are smart, but they’re only human and figuring out what thousands or even millions of people will like is a gargantuan task. This is where AI such as predictive analytics comes in handy.
While sifting through the myriad preferences of thousands or millions of people is impossible for a person, it’s very practical for a computer. Software can be and is used to parse through transactional, web and email data, and apply advanced machine learning to that data in order to discern the preferences of each individual customer.
Armed with this insight, the marketing team can better determine which audiences should be exposed to which products and content. Creatives can then dedicate their energy to the topics and products that are most likely to resonate.
2. It creates more opportunities for work–and more impactful work
AI applications such as predictive analytics enables marketers to segment audiences into more refined and rational segments based on customer interest than do traditional demographic (i.e. gender-based) segmentation or RFM segmentation. Often this results in more segments than traditional methods would produce. This gives you the option of putting together creative for each segment, which means more work for creative professionals rather than less. Moreover, because the segments are drawn according to customer interest, each campaign, ad and promotion is likely to be more effective than it would be otherwise.
Granted, it may not be practical to create custom copy, graphics and photography for a dozen plus segments, but the point remains that machine learning — at least in the context of marketing — can help create rather than diminish opportunities for work.
3. It helps you know your audience in more relevant ways
Personalization and predictive analytics systems draw on AI to predict and adapt to customer behavior. You can think of these technologies as performing a similar role in digital channels to that played by store keepers, salespeople, personal shoppers and in-store associates in physical stores: they get to know the customer’s habits and preferences.
Coherent Path solutions, for example, discern not only customer preferences but the fluctuating nature of these preferences. People have tastes, but these tastes have a rhythm to them and change over time. Think about food you like. You may like pizza, but you probably don’t want to eat pizza everyday. On some days you’ll prefer salad; on others, fish tacos. The same is true for consumer purchases. Indeed, this is why many recommender engines and retargeted ads annoy people: after purchasing an item, you’re probably not terribly inclined to view a dozen more ads of the same item and purchase it again. Most people don’t buy a dozen road bikes in two weeks, for example.
An in-store professional can probably get to know the rhythm and tastes of a dozen or so customers, but a marketer in the corporate office is not going to be able to do so for thousands of customers. Again, software that uses machine learning can accomplish this task quickly and efficiently, allowing the marketer to be more effective at his or her job.
People like Stephen Hawking and Elon Musk are smart and well-informed, to say the least. So, if they think AI poses certain risks, we’re not going to argue with them. However, in our experience delivering machine learning marketing applications to retailers, we’ve found that this technology rather than threaten marketers and creative professionals, actually enables them to do their job more effectively and easily than before.