Getting Smart About Segments - Data-Driven Buyer's Guide

If we think about their origins in direct marketing, "marketing segments" seem an outdated methodology. Yet, they are still in play, with publishers and data solutions providers selling them across channels as targeting options. They often tout or over-sell these simplistic, static audience segmentations based on demographics and behavioral data as state-of-the-art dynamic audience propensity models.

But, that’s often not what they are.

Looking under the hood, these segments don’t incorporate preferences unique to a particular user—things that are possible to gauge today. So, it’s less likely that they target a marketer’s unique consumer. If your competitor is buying the same generic data segment(s)—they are targeting the same audience pool. Rather than that segment representing your authentic target audience, it may be a general one that is prevalent and popular in the industry. Your product—and its user propensities and behaviors—are left by the wayside, while you mirror the tactics of your competitors.



To get more adept, there are considerations and questions to ask when meeting with planning teams, publishers and data tech solution providers on data targeting options.

Look beyond the Targeting Norm
We’ve been using demographics to establish targeting parameters for digital marketing since the dawn of digital marketing. Then, layering on standard behavioral targeting—content preferences or browsing data, we have established a slightly better norm. But, from the vantage point of the enlightened data driven marketer, lifestyle, mindset, behaviors and intentions together matter so much more than basic demographic or geographic facts that make up that norm dubbed a segment.

Marketers should up the ante on their methods for identifying and targeting audiences—and what they expect. New data signals allow us to do so much more than we could before. At every touch, the consumer is relaying a new "tell" and revealing their propensity for action that can be optimized and in real-time. You want partners prepared to integrate these signals and to interpret them.

Here are a few questions to address, to strengthen your position, when engaging the agency or the potential provider of data tech solutions:

1. How are the segments developed? It’s important to know exactly which attributes are considered, how they are clustered and rolled up to constitute any given segment. If the roll-up is too simplistic, it’s rational to ask: do we lose some of the valuable data and specific attributes that live inside these clusters? Doesn’t the simplistic audience segment become more of a community segment? And, how does that kind of segment reflect my target customer? Concrete “attributes” are essential to your marketing message. We need the specifics. Otherwise we are just targeting a broad swath—seemingly inclined at a glance but not necessarily interested in your product, based on their real values.  

2. How can we get volume, based on the specific attributes I value? As a provider talks to you about a broad segment opportunity, you would be wise to ask: if I am targeting the cluster bereft of the detail of the attributes important to me, how can we know how many people with those more valuable specific attributes are within the total segment? How do we pull those people out and scale? If you already know the data you desire is available and potentially living inside the modeled segment, why would you drop it? You need to find a solution that surfaces and leverages it.

3. Is this targeting or just high-level guesswork? If a marketer is sticking with community based segments, can we really use these models to target—and/or is this segment a proxy based on a high-level predictive tool?  If the latter, what other data can we integrate to strengthen our ability to model toward our outcome and target volume?

4. How can I personalize? I’m planning to personalize creative—visuals, messaging and call to action—based on individual profiles within the model. Won’t we need individual vs. community signals to be able to do that?

5. How do we set the stage for truly dynamic audience optimization? Typically, segments or models are not as frequently or dynamically optimized in real-time as they could be. Once per day, let alone weekly or monthly, is an outdated optimization approach—given what today’s systems are capable of handling. With a focus on continually improved efficiency and growing ROI, you need to know how a partner will be tuning personalized creative and our execution in real, real-time. So—ask.

With something far more sophisticated than the generalized audience segments now possible, it’s essential to play smart—as you evaluate what publishers and data tech solution providers propose. The key is to have the opportunity to model and target according to the attributes you really value. Otherwise, if you accept commonly sold and used audience segments, you are essentially buying an off-the-shelf commodity—and the same one your competitor may be too.

Next story loading loading..