From text and images to videos and avatars, artificial intelligence has become the superpower that allows organizations to mass produce content and assemble elements in endless combinations. But the ability to quickly launch multiple campaigns is a meaningless victory if messaging and marketing fail to treat people as individuals.
Personalization unlocks the potential for growth, but delivering the right message to the right customer in the right context today relies largely on rigid, rule-based approaches that obey logic in an age where consumer behavior doesn’t. has never been so changeable. “People are unpredictable, and their journeys and decisions are not linear. » Sam Richardsoncustomer engagement consultant at Twiliosaid Marketing Tech in an interview. According to her, it is “time to abandon the old methods and models of customer demographic segmentation.”
The one from Twilio State of Personalization Report 2023 presents the data to save the information. Nearly half (48%) of marketers surveyed in the report question the value of traditional customer segmentation. Instead, they are shifting their focus and budget toward more fluid approaches to optimize how they segment and serve their customers. (The report, detailed in this article by Forbes contributor Daniel Newman, highlights the central role of personalization in a strategy to build 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 complexity. “This is where first-party data is pure gold, but not necessarily the last word. » Alon Rivela growth marketer known for his out-of-the-box thinking, told me in an interview.
He explains how he combined user data and behavioral insights based on consumer activity online and in apps to create compelling messages that activate and motivate at scale. This approach has been a common thread throughout his experience at various brands, including Soothe, an on-demand wellness marketplace, Outcomes4Me, a patient empowerment platform that helps cancer patients manage their care proactively, 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 trust marketing to understand the nuances of their context, Rivel says. “A disconnect doesn’t just create friction; this can generate churn. To avoid both, it’s up to marketing to evolve in sync with where the customer is in their journey. Let’s take the example of a user who takes the plunge and pays for a premium feature in a subscription app. Naturally, a marketer may read this as a green light to suggest an upgrade to a one-year premium subscription – and there are engagement models that will support this, says Rivel. But even machine learning has a lot to learn.
“While the action indicates high intent, it does not automatically trigger immediate delivery of a limited-time offer,” says Rivel. “An effective message aligns with a customer’s behavior up to that point, focusing on their propensity to engage positively with the offer presented to them. In this scenario, he says, bombarding users with features, deals and discounts is a failure unless their actions clearly signal a match. “The future marketer follows the customer’s lead.”
Customer-Driven Segmentation Gets Personal
A customer-focused segmentation approach is at the heart of a bottom-up approach championed by Aampé, a company that leverages AI to deliver “messaging-based” personalization at scale. It does this by ingesting and modeling an application’s event stream (the continuous flow of real-time information generated by user interactions, system events, or other activities within the application) to to detect meaningful patterns in user engagement. The result 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 increase retention.
“While businesses are freed by AI to adapt their marketing to the speed of change, they are limited by rigid segmentation constructs and prone to oversimplification.”
This is a significant departure from popular rule-based approaches that group people based on observations about what they do (or don’t do) on a daily, weekly, or monthly basis. “This thinking constitutes a remarkably anachronistic approach to personalization and is why we choose to focus on discovering and understanding the attributes that drive real, measurable customer engagement. » Paul Meinshausenthe co-founder and CEO of Aampe told me in an interview.
This is new territory for marketers and a new category of technology developed by Aampe management that combines talents in data science, data engineering, neuroscience, anthropology and product development. “Personalization must view customers through a living and learning lens that views them as individuals,” says Meinshausen. “To make this possible, our toolset must evolve, and that’s exactly what AI is driving, an evolution of software.” In this scenario, Aampe offers a “reinforcement learning infrastructure,” ensuring that organizations do not become stale or static in their communications.
Interesting opportunities among existing customers
Along with this increased focus on changing consumer behavior, message personalization requires continuous experimentation and iteration of content, context, timing, tone and sentiment. The combinations for personalizing messages are endless, but the patience, time and resources of marketers are not. This is where AI shines, automating the manual work of content creation, experimental design, and conversion tracking.
Freed from the tedious work of generating and testing messages to find the right solution, and freed from preconceptions about which customer segments are most likely to respond to which messages, marketers will discover the inspiration and insights needed to busting the biggest marketing myths.
Consider the timing and frequency of messages. People don’t follow rigid routines. Yet most marketing activities follow flowcharts and rules-based frameworks that make general assumptions about best practices and the best time to engage customers. It is not surprising that messages are often inserted into specific times of the day (for example early in the morning, from 7 a.m. to 9 a.m.; midday, during the lunch break from 12 p.m. to 2 p.m.; and early evening, from 6:30 p.m. to 8:30 p.m.). And because consumers are inundated with notifications during these times, notifications arriving during these times tend to add to the noise, not exceed it. This rigid thinking and planning can limit marketing’s ability to extract new value (and additional revenue) from existing customers.
That’s what one of Asia’s super apps discovered when it took a closer look at its system’s rule to only send in-app notifications at 12 p.m. and 6 p.m., Meinhausen says. Sending messages during conventional meals made good business sense, but it turned out not to be the best way to drive additional orders or increase revenue.
“Programmatic experimentation with Aampe revealed that messages delivered until 11 p.m. attracted the attention of a small but stable cohort of app users,” recalls Meinhausen. Data analysis revealed a new underserved customer segment: people who regularly worked late and were hungry for energy-boosting late-night and evening snacks. Rethinking campaigns to cater to this highly valuable and loyal segment, ignored by standard CRM software, allowed the app company to generate a “24-38% increase in incremental new transactions from this specific population of users », he adds.
Examples like these reveal a fatal flaw in approaches biased by rules-based systems. While AI allows businesses to adapt their marketing to the pace of change, they are limited by rigid segmentation structures and prone to oversimplification. 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 message content), the “when” (the best message timing and frequency), and the “why” (the consumer context and state of needs), building on what we Really doing and valuing, offer a compound path to profits.