A previous blog on the evolution of audience segmentation focused on the benefits of automated segmentation through machine learning, specifically how unsupervised models discover data correlations that reveal something important about an audience, beyond the capability of an operational marketer.
Dynamic segmentation through machine learning solves another challenge that marketers have when it comes to audience selection, which is that audiences – i.e., customers – change faster than marketers can keep up. Digital-first customer journeys are becoming more complex by the day, with customers randomly moving between digital and physical channels, yet still maintaining an expectation that the brands they engage with demonstrate a personal understanding.
The challenge stems from the fact that a list-based approach to audience segmentation, a standard practice for many marketing organizations, does not keep pace with a dynamic, omnichannel customer journey. Selecting an audience through a static list may suffice for a “batch and blast” outbound campaign restricted to a single channel, but the list does not account for the fact that customers freely move between channels. It will also become static the moment a change occurs that would otherwise add or remove a customer from the list – the customer makes a purchase, signs up for a loyalty program, downloads a manual, etc.
Staying In Step with a Customer
A rules-based, dynamic approach to audience segmentation in the rg1 platform, by contrast, by accounting for real-time pivots, guarantees that a brand keeps pace with a customer as the customer moves through a customer journey.
Unbound by channel, rules-based machine learning finds audience similarities and tests segments on the fly. The algorithm trains on real-time data that is presented to the model, and the algorithm – not the marketer – dynamically selects the optimal audience at the time of selection, on demand, up to the last possible moment before campaign execution. It’s like a living, breathing model that changes in step with the customer.
Because the model analyzes the composition of an audience on the fly, it becomes re-usable. Unlike static lists that decay as soon as they are created, a rules-based approach enables organizations to create a segment that can be universally applied across any engagement channel. Information is collected in real time, and new data about a customer might move the customer to a different channel. The customer who signs up for the loyalty program, for example, may no longer be qualified for the segment queued for an email offer, and is instead moved to a segment of high-value customers who receive a biweekly newsletter with a VIP redemption code.
Dynamic audience segmentation is a key feature that allows a company to look at customers in their entirety across the breadth of a customer journey vs. a view constrained by interactions on a single channel.
In rg1, customers fall into and out of a segment based on their latest activities as they interact with the company, across all channels. It is a key distinction with a list-based approach, because it allows a brand to engage customers with up-to-the-moment relevance and deliver hyper-personalized experiences that reflect a customer’s precise customer journey status.
Basing an audience segment on static customer profiles, by contrast, fails to provide the cross-channel insight required to keep pace with a customer on a typical omnichannel journey. Knowing how a customer behaves only on a single channel is of little help when a brand is reaching out on another channel. By segmenting audiences based on customer insights across many channels and touchpoints in real time allows a brand to use a customer’s behaviors in one channel to inform how the brand will engage with the customer on another channel.
In this fashion, segments are created not by using mostly static parameters such as a customer’s age bracket, location or income, but far more nuanced (read: relevant) attributes as determined by a machine learning model. Think of your own experiences as a customer; when a brand reaches out to you only because you fall into a certain age demographic, does it feel impersonal? Your age may have little bearing on movement through an omnichannel customer journey. Dynamic audience segmentation in the rg1 platform gets at the heart of what’s important to a customer at any given moment in time, allowing a brand to consistently serve up relevant interactions in real time and on any channel.