When It Comes to Targeting, Artificial Intelligence Isn't a Panacea
It’s difficult to overstate the digital marketer’s embrace of artificial intelligence. Demand-side platforms leverage machine learning algorithms to select the right message to present to the right consumer. AI is used to optimize Google Ads bidding strategies.
And unsupervised machine learning is deployed to look across the universe of consumers, identify the behaviors that indicate someone is in the market for shoes or a BMW, and target that person with brand messages.
It also allows marketers to segment consumers based on attributes never previously considered.
No Substitute for Knowing Your Customer
These are powerful reasons why the market should invest in AI. But as great as these applications are, we must take care not to view AI as a panacea for campaign targeting for an important reason: It can predict behavior, but not motivation. And when it comes to finding long-term and profitable customers, motivation is everything.
Let me explain. AI excels at identifying consumers who are actively in the market for a product or service, but it can’t determine if a person will be a one-and-done consumer or someone who is likely to have a high lifetime value for a brand. This is an important distinction, given that most brands rely on advertising to build their upper sales funnel (and by extension, secure their futures).
All sales are good, of course. But if it costs $50 to acquire a new customer, wouldn’t you want to ensure every new customer had a high lifetime value? Fifty dollars for a customer who will spend $500 annually over a 10-year span is a great investment. But if a consumer makes a one-time $30 or $40 purchase, it’s a colossal waste, compounded by the fact that this investment could have been better spent elsewhere.
The Truth About AI
There are those who say that AI can identify and target consumers who have a high lifetime value for a brand, but we’re not quite there yet. The challenge is that targeting is largely based on a combination of demographics, psycho-demographics and behavior. But to really target consumers with the potential for high lifetime value, one needs to understand and target against the motivations for making a purchase.
Let’s break this down. Why does a female consumer purchase a pair of red pumps? Maybe she loves fashion and they’re an item she simply can’t resist. If that’s the case, she’s a consumer who is likely to have a high lifetime value for a shoe brand.
Or maybe, she purchased an expensive pair of red pumps because she’s in a wedding party and they’re what the bride wants her to wear. It may be that those pumps are so far from what she would typically buy, which may be whatever is comfortable, practical and affordable. She may be highly satisfied with the shoe quality, price she paid, packaging they came in and service she received during the sale, but she’d still prefer never to buy another pair of pumps again.
So, on one side, we have a consumer who adores fashion, and has a deep emotional connection to it. Buying new clothes makes her feel creative and empowered and helps her stand out from the crowd. On the other side, we have a consumer who is potentially disaffected by the category. She buys clothes when her old ones wear out or in response to a need, such as a new job with a specific dress code. These two women have two different emotional approaches to clothing, and those emotions drive their behavior and, ultimately, their value to fashion brands.
Emotionally Motivated Conversions
The financial impact of emotional connection is both eye-opening and compelling. We compared consumers who are emotionally connected to a product or category to those who rate themselves satisfied, and found the former:
- Spend up to two times, or more, with their preferred retailers
- Have a 306 percent higher lifetime value
- Stay with a brand for an average of 5.1 years vs. 3.4 years
- Recommend brands at much higher rates: 71 vs. 45 percent
When AI Matters
The upshot is that some people, for reasons they can’t articulate, are just hard-wired to gravitate to a particular category. Of course, psychologists have long recognized this, and it’s worth taking their observations to heart. As marketers, we want 100 percent of our ad spend focused on people who are hard-wired to spend a greater share of their wallet in our categories. Once we identify who they are, no doubt the AI-infused marketing execution ecosystem will excel at reaching them.
So deploying AI for the purpose of identifying who is in the market for a particular item is good, but using it to predict who is likely to purchase over and over with your brand — and feel great about those purchases — is what will really change the game for marketers.
Related story: 5 Tips to Make Machine Learning Work for Brands