« View all blog posts

Behavioral segmentation data outperforms demographic data

Demographic-based segments have been marketing's bread and butter for many years, and they're pretty straightforward. Create a persona, choose some attributes that define the persona, then formulate some rules to target users with those attributes. Et voila! You've got a segment.

While by while the approach is straightforward, it turns out that you're not likely to get much mileage out of it.

How do we know? Well, we took our own data and used it to figure out the subsets of users most likely to interact with a campaign. And when we did so? It was clear that behavioral data was leaps and bounds more predictive of engagement than demographic data.

If you're not already incorporating behavioral data into your workflow, you definitely should be. The only problem is that it can be difficult.

Taking behavioral data by the horns

Why is this so hard?

If you want to incorporate your behavioral data into your workflow, you're often confronted with it head on and are immediately required to make the following choices:

  1. Choose your fields: Identify fields in your customer database that identify user behavior, like purchases, email opens, site visits, social shares, etc.
  2. Choose your thresholds: Determine the value of the field you chose that will best identify the users you're trying to target, like more than one purchase, two email opens or greater, etc.

Constructing rules for execution can be a fragile process. Let's say you're trying to run an upsell campaign for users who are "mildly active" who you'd like to convert into "loyal" users.

Have your "mildly active" users opened one email? Or two? Or twenty? Or better yet, maybe they clicked on a link in the email.

If that’s the case, which field do you use to identify the users you're trying to target: Opens or clicks? Or do you reconcile both somehow?

As you can see, the questions pile up quickly. And the efficacy of your execution depends on if you happened to choose fields and thresholds that encapsulate the behavior you’re trying to target.

A better way to use behavioral data

As we've discussed before, there are a lot of aspects to user behavior, and Lytics' behavioral scoring helps to summarize them into dimensions with which you can get your hands dirty.

Let's go back to the upsell campaign we mentioned earlier. First, we'll want to identify your users who are "mildly active," for which we can either:

  1. Use the Active segment that comes pre-built in Lytics,

  2. Create a more behaviorally descriptive segment using raw Lytics scores like Momentum, Quantity, and Intensity.

Then we can segment further on users who aren't currently in your loyalty program.

Bingo! You've now targeted users who meet specific behavioral conditions in just two steps. Sync these behavior-based segments to your email service provider and start a nurture campaign to influence these customers towards becoming advocates of your brand, as an example.


Not only is it easier to use your behavioral data using tools in your Lytics toolbelt, but it helps you be an incredibly effective marketer.

When we run an analysis of all our data, including Lytics' behavioral scoring data, to identify users most likely to engage with a campaign, it's easy to see that users targeted with behavioral scores are 20x as likely to interact with your campaign than users targeted with demographic attributes alone.

Which means whatever your workflow may be, it's vital that you're somehow incorporating behavioral data to target your engagement with your users.

And if you're ready to take that step? We're here to help. Contact us today to schedule a demo.