From text and images to video and avatars, artificial intelligence has become the superpower that allows organizations to produce content in bulk and assemble elements in infinite combinations. But the ability to quickly churn out multiple campaigns is an empty victory if messaging and marketing fail to treat people as individuals.

Personalization unlocks growth potential, but delivering the right message to the right customer in the right context largely relies today on rigid rules-based approaches that obey logic at a time when consumer behavior has never been more changeable. “People are unpredictable, and their journeys and decisions are non-linear,” Sam Richardson, customer engagement consultant at Twilio, told Marketing Tech in an interview. In her view, it’s “time to ditch the old demographic customer segmentation methods and models.”

Twilio’s State of Personalization Report 2023 lays out the data to back up the insight. Nearly half (48%) of marketers surveyed in the report question the value of traditional customer segmentation. Instead, they are shifting the focus and budget to more fluid approaches to optimizing how they segment and serve their customers. (The report, detailed in this article by fellow Forbes contributor Daniel Newman, underlines the pivotal role of personalization in a strategy to drive customer loyalty and increase LTV.)

Balancing signals for tangible benefit

Delivering a VIP experience to a mass audience requires organizations to understand customers in all their rich complexity. “This is where first-party data is pure gold, but not necessarily the last word,” Alon Rivel, a growth marketer recognized for his out-of-the-box thinking, told me in an interview.

He shares how he combined user data and behavioral insights based on consumer activity online and in-app to craft compelling messaging that activates and motivates at scale. This approach runs like a thread throughout his time at diverse brands, including Soothe, an on-demand wellness marketplace, Outcomes4Me, a patient empowerment platform that helps cancer patients proactively manage their care, and Lose It!, an app that ensures people set and stick to their weight loss goals.

Customers rely on these services for personal assistance, and they trust marketing to understand the nuances in their context, Rivel says. “A disconnect doesn’t just create friction; it can drive churn.” To avoid both, it’s up to marketing to move in lock-step with where the customer is in their journey. Take the example of a user who makes the leap to pay for a premium feature in a subscription app. Naturally, a marketer may read that as the go-ahead to suggest an upgrade to a year-long premium subscription—and there are engagement models that will back that up, Rivel says. But even machine learning has a lot to learn.

“While the action indicates high intent, it doesn’t inherently trigger the immediate delivery of a limited-time offer,” Rivel explains. “Effective messaging aligns with a customer’s behavior up to that moment, focusing on their propensity to engage positively with the offer they are presented.” In this scenario, he says, bombarding users with features, offers, and discounts is a fail, unless their actions distinctly signal a match. “The future marketer follows the lead of the customer.”

Customer-led segmentation gets personal

A customer-led approach to segmentation is at the core of a bottoms-up approach championed by Aampe, a company that harnesses AI to deliver “messaging-led” personalization at scale. It does this by ingesting and modeling an app’s event stream—the continuous flow of real-time information generated by user interactions, system events, or other activities within the app—to detect meaningful patterns in user engagement. The output is a deep understanding of the customer based on how, when and why they interact with messaging, which can inform campaigns to increase conversions and grow retention.

“While companies are liberated by AI to tailor marketing at the speed of change, they are limited by segmentation constructs that are rigid and prone to over-simplification.”

This is a significant departure from popular rules-based approaches that bucket people according to observations about what they do—or don’t do—on a daily, weekly or monthly basis. “That thinking is a remarkably anachronistic approach to personalization and why we choose to focus on uncovering and understanding the attributes that drive real and measurable customer engagement,” Paul Meinshausen, Aampe co-founder and CEO, told me in an interview.

It’s new territory for marketers and a new category of technology developed by Aampe leadership that combines talents in data science, data engineering, neuroscience, anthropology and product development. “Personalization has to view customers through a living and learning lens that frames them as individuals,” Meinshausen explains. “To make that possible, our toolset has to evolve, and that’s exactly what AI is driving, an evolution in software.” In this scenario, Aampe offers a “reinforcement learning infrastructure,” ensuring organizations don’t become stale or static in their communications.

Exciting opportunities among existing customers

Along with this sharper focus on evolving consumer behavior, messaging personalization demands ongoing experimentation and iteration in content, context, timing, tone and sentiment. The combinations to tailor individualized messaging are infinite—but marketers’ patience, time and resources are not. This is where AI shines, automating the manual work of content creation, experimental design and conversion tracking.

Freed from the drudge work of generating and testing messages to find the fit—and liberated from preconceived notions about which customer segments are highly likely to respond to which messages—marketers will uncover the inspiration and insights to bust marketing’s biggest myths.

Take message timing and frequency. People don’t follow rigid routines. Yet, most marketing follows rules-based flow charts and frameworks that make sweeping assumptions about best practices and the best time to engage customers. Little wonder messaging is often squeezed into specific moments of the day (for example, early morning, from 7 AM to 9 AM; mid-day, during the lunch break from 12 PM to 2 PM; and early evening, from 6:30 PM to 8:30 PM). And, because consumers are inundated with notifications during these periods, notifications arriving during these times tend to add to the noise, not rise above it. This rigid thinking and scheduling can limit marketing’s ability to extract new value—and incremental revenue—from existing customers.

That’s what one of the Asian super apps discovered when they took a closer look at their system rule to only send in-app notifications at noon and at 6 pm, Meinhausen says. Messaging during conventional mealtimes made business sense, but it turned out not to be the best way to drive additional orders or grow revenues.

“Programmatically experimenting with Aampe revealed messaging delivered as late as 11 pm was getting the attention of a small but stable cohort of app users,” Meinhausen recalls. Digging into the data further revealed a new and underserved customer segment: people who regularly worked late and were hungry for energizing late evening and night snacks. Rethinking campaigns to cater to this highly valuable and loyal segment, ignored by the standard CRM software, allowed the app company to drive a “24-38% increase in incremental new transactions from that specific population of users,” he adds.

Examples like these expose a fatal flaw in approaches biased by rules-based systems. While companies are liberated by AI to tailor marketing at the speed of change, they are limited by segmentation constructs that are rigid and prone to over-simplification. It’s a dynamic that makes winning with AI-driven personalization a bit of an obstacle course. Fortunately, approaches that connect the dots between the “who” (the correct audience segment), the “what” (the appropriate messaging content), the “when” (the best messaging timing and frequency) and the “why” (the relevant consumer context and need-state), drawing on what we really do and value, offer a compounding pathway to profits.

Get Film Insurance Today
Author: