I remember the promise years ago in which the prevailing buzzword was "one-to-one marketing." The dirty little secret everyone in the room knew back then was that the term itself was a lie. Marketing could never be one-to-one at that time because technology did not support it. However the name "one to a targeted group that looked like one" would not really have sold anyone on the premise. Neither the data nor the technology existed.
Fast forward a couple of decades. Today we see the explosion of digital, big data, and the warehouses and technology required to support it, as well as the enhanced analytic capabilities required to infer from it. This synthesis of data, technology, application capability, and human intelligence has been organized into an integrated technology stack that can be deployed and adapted to deliver on the promise of what was then called "one-to-one marketing."
To address a segment of one, we trace and evaluate a customer’s behavior based on his or her digital footprint, take the appropriate action based on that behavior, and create targeted messages just for him/her — and for, conceivably, thousands of other individuals, one at a time. This promise is not without its pitfalls in execution. Realizing satisfactory results in this new world requires the use of many advanced systems: heavy-duty analytics, highly sophisticated targeting methods, complex campaign decision trees. You must master these to support today’s customer-initiated engagement model.This carries greater complexity and with it a different expense model than existed in the “old days.” Controlling costs and realizing strong marketing ROI requires test marketing. It always has. The difference today lies in the granularity of the testing required and, therefore, the number of test cases that must be developed to arrive at a representative sample.
To illustrate, in the B2B world, marketers developed segments based on attributes such as employee size, sales volume, job role, year business started, and industry; then they would test a limited number of messages to samples taken from these wide audiences. Today we use the same criteria to build audiences but treat these audiences as smaller populations from which we sub-select to test the effect of small variances in content — subject line, placement, message — as applied to individuals, true “audiences of one.” This process also relies on the need for more content to be available to these micro segments.
The point of greatest cost control has therefore shifted. Before, the greatest cost was in creating, deploying, and maintaining multiple creative variations. Today the cost is in managing a greatly increased number of targeted “segments of one.” So today, test marketing involves a greater number of smaller variations off of the control piece, distributed over a greater number of recipients. Analyzing the results involves analytics’ searching for common criteria among both responders and non-responders to specific, controlled variations in content.
Test marketing today involves more moving parts than ever before. Fortunately we have the tools, data, and capabilities to control these moving parts and deliver actionable guidance to marketers by carefully controlling small variations and allowing the data to define individuals’ propensity to respond and engage.
This is just the first part in the changing landscape of one-to-one communications. Please stay tuned for my next blog in the series, titled "True multi-wave campaign communications."