It’s no secret that modern digital marketers need more than just declarative data to build quality, personalized campaigns. They’ll need more than a snapshot of clicks, opens, retweets, and other signals to build quality, personalized relationships with customers.
Personalization needs to be rooted in a firm understanding of how a user is interacting with your brand – a.k.a. behavioral, or engagement, data.
What, then, is preventing marketers from fully utilizing their behavioral data?
Let’s try answering that question with another question: Do you call your mother?
The answer is probably more nuanced than a simple “yes” or “no” can provide. Maybe you call her daily or weekly, just to say hello. Maybe you call her monthly, but the conversation is deeper than a typical weekly check-in. Maybe you just call on holidays. Or, maybe it’s been years since you’ve called.
In data science, we’d call the answer to that question a flag field, which reduces a complicated answer to a simple one – yes or no. Black, or white.
A lot of people apply the same simplification of behavioral data down to a single flag field, or even a set of flag fields, which makes behavioral data become declarative. But, human behavior is seldom dichotomous – it’s gradual, nuanced and contextual. It’s multi-dimensional.
In other words, we’re going to need more colors.
Lytics provides cross-channel, multidimensional behavioral analysis out of the box by providing a suite of behavioral scores that identify specific aspects, or dimensions, of user behavior. Our process lets us see users in a variety of shades of their behavior, allowing us to group them in ways for marketers to improve the level of personalization in their campaigns.
Here’s a list of Lytics behavioral scores:
Consider two different theoretical users visiting an e-commerce retailer. Both visitors have 60 pageviews and one purchase in the month of May.
Allie is a long-time user who likes to peruse the site, often during lunch breaks.
Michael had a hard time finding a birthday present for a friend during his first visit to the site, but eventually found something that he liked.
By a lot of behavioral metrics, Allie and Michael look the same, but the nature of their engagement with the retailer is drastically different. Here’s how they would look by their Lytics behavioral scores:
The context around the conversation that Allie needs is much different than what Michael needs. Michael still needs more nurturing, and we’d characterize him as a binge user. Allie is well-aware of prices and products, and always tries to find the best deal that she can. We’d characterize her as a peruser.
We now know that Allie and Michael are both recent purchasers, but their behaviors are completely different. Knowing this, would you send them the same marketing messages? This seems counter to actual personalization.
The fact that Allie is the type to know what she wants and is very price sensitive, we’ll want to respond to her habits of heavily perusing and rarely taking action. Therefore, we can hypothesize that she is more likely to respond to marketing campaigns that contain discounts or incentives to purchase.
Michael probably needs additional product recommendations based on his interest since he doesn’t know what he wants. We can put him into a retargeting campaign that will continue to drip-feed relevant products of interest to him until he finally decides to make another purchase.
Allie and Michael are two of many behavioral archetypes that businesses should classify their marketing segmentation towards. Lytics makes this kind of segmentation by behavior easy to do with your existing marketing data.
Lytics pre-packages behavioral scores into useful, consumable, and actionable segments. We call these our Behavioral Segments.
These pre-built segments include:
Our Attribute Segments break down behaviors between each of our segments, letting marketers really see the motivations of how users traverse between each segment.
These Attribute Segments are:
You can start using Lytics behavioral scores and pre-defined segments shortly after account creation. Explore different score combinations to identify behavioral clusters unique to your audience, or let us handle the segmentation so you can focus on execution.
However you use them, our behavioral scores probably won’t replace your existing segmentation. But they will definitely make it better.