Once your data sources are connected and models are defined, the next step is to use this structured data to build powerful audience segments. In NotifyVisitors, segmentation allows you to group users based on demographic attributes, shared behaviors or actions, enabling you to deliver personalized experiences and highly targeted campaigns.

With query-based segmentation, you can directly leverage external data models created from your data lake. This gives you the flexibility to build dynamic, rule-based segments that always reflect the most current data from your connected sources.

What is a model-based segment?

A model-based segment is a dynamic audience group created from the data stored in your connected models. These segments are built upon user or event models and can be further customized by applying attribute-based filters. You can filter users based on demographic details like age, city, account status, and plan type, or behavioral events such as purchases, logins, cart abandonments, or completed transactions. Additionally, any custom column in your model, such as subscription tier, wallet balance, or cancellation reason, can also be used to refine the segment.

Once created, these segments can be activated across the NotifyVisitors platform, allowing you to trigger personalized campaigns across channels like email, SMS, WhatsApp, push notifications, and more. Whether your goal is re-engagement, upselling, or retention, model-based segments provide the precision needed to make your campaigns more effective.

How to create a model-based segment

To create a model-based segment, start by navigating to the Segments section in the NotifyVisitors dashboard and select the Segments tab. Click on Create Segment and choose Query-Based Segment as the type.

Begin by selecting the data source from which you’ve imported your models. All your connected sources will appear in the dropdown list. After selecting the source, click the Continue button. This opens the segment builder, where you can construct rules using your connected models.

Segment builder

In the segment builder, first select the model you want to base your segment on. If needed, you can also add multiple models to the query to build cross-model logic. Each model acts as a separate rule block. Within each rule, you can click Add Attribute to apply filters based on specific columns, whether those relate to user profiles or recorded events.

For instance, imagine you're a fintech company looking to target users who haven’t transacted in the past 60 days, but who also have a high wallet balance and belong to the "Premium" membership tier. In this case, you could use one model that tracks user profile data, including wallet balance and membership tier, and another event model that logs financial transactions. You would then apply a condition on the event model to exclude users who’ve made a payment in the last 60 days and filter the user model to include only those with a wallet balance over ₹10,000 and a membership tier marked as “Premium”. This cross-model logic ensures you're reaching a highly relevant audience.

Once your rule blocks are defined, choose how they should be evaluated together. If you want users to qualify if they meet at least one of the conditions, select “Any rule can satisfy”. If users must meet all the conditions from each model you’ve included, choose “All rules must satisfy”. In our example above, since we want users to meet all of the conditions, we’ve selected the “All rules must satisfy” option.

Before finalizing, you have the option to enable the Precomputed setting. When this is turned on, NotifyVisitors will automatically update the segment based on the latest synced data just before the campaign is sent. This ensures your audience is always up to date without requiring manual refreshes.

After reviewing your conditions, click Create, give your segment a name, select a label (optional), and then click save. With this your segment will be saved as draft. If you want publish right away, simply click the Publish Segment button. 

The segment will be added to your segments list and will be ready to use in campaigns.

Segment list view and management

In the segment list view, you’ll find all previously created segments organized by name (along with campaign ID, last updated date, and creation date), status, user count, and more. Each segment also carries a label indicating its type, such as “Query-Based Segment,” “Journey Segment,” or “Past Behavior Segment.” If you want to filter only query-based segments, simply use the Segment Type dropdown and select Query-Based Segment. The list will automatically refresh to show only Query-Based Segments.

Each row in the segment list also includes an action menu (represented by three dots) that offers additional controls. You can view detailed analytics for that segment, see which campaigns it has been used in, access user logs, duplicate it, delete it, or manually trigger an update. Editing is permitted only for segments that are still in draft mode. Once a segment has been published, it becomes locked from further edits to preserve campaign accuracy and consistency.

Why use model-based segmentation?

The Model-Based Segmentation method removes dependency upon moving your data to NotifyVisitors. Since you’re working directly with data from your own data lake, including user attributes and events, you get complete flexibility in how you define your audience. You can combine data across tables, filter on specific attributes, and create segments that reflect user behavior. This enables granular and precise audience targeting.

Whether you’re targeting dormant high-spenders, recent drop-offs, or users who meet complex behavioral and demographic combinations, model-based segmentation gives you the tools to do it with confidence.