Create Segment with Co-Marketer
Learn how to create NLP based segments with Netcore's Co-marketer.
Overview
Co-marketer segmentation lets you create audience segments by simply describing the audience in plain English, rather than manually building complex rule sets. You can type what you want. For example, “Users who installed the app in the last 7 days and added items to cart but didn’t purchase” and the system automatically converts it into structured segmentation rules inside the platform.
This feature is designed to:
- Reduce the time and effort required to build segments.
- Make segmentation accessible to non‑technical users.
- Minimize dependency on deep data and attribute knowledge.
Key Benefits
- Create segments faster: Go from idea to usable segment in seconds
- Use natural language: No need to understand complex rule builders upfront
- Lower learning curve: Ideal for new users and non‑technical teams
- Full control after creation: Review, edit, and refine the generated rules manually
Important Points to Remember
- Co-marketer segments are not available while editing an existing segment.
- You can submit only one request at a time.
- There are no follow‑up questions from the system.
- The system generates one complete segment
- After that, you can refine the segment manually in the builder
Create Segment with Co-marketer
Create Segment with Co-marketer
Follow the steps given below to create a segment with Co-marketer:
- Log in to the Netcore CE dashboard and navigate to Audience > Segments
- Click CREATE on the top right corner.
- The Create New Segment page opens. Here, the Co‑marketer panel is visible by default on the right side of the screen.
- In the Co‑marketer input box, either:
- Type your audience description in English, or click one of the three suggested prompts.
Example prompts
- Generate a segment with users who installed the app in the last 14 days.
- Create a segment with users who added to cart but didn’t purchase in the last week.
- Find inactive users who haven’t opened the app in 30 days.
- After submission the card displays:
- A summary of the interpreted audience & An estimated Reachable Users count
This allows you to validate whether the system understood your intent quickly.
- Click View on the summary card to automatically populate the segment builder. The system has now converted your request into structured rules. All conditions now appear in the standard dropdown‑based segmentation UI.
- After validating your rules, click Save
Your Co-marketer segment is now available like any other segment and can be used across campaigns.
Example queries and expected results
Refer to the given table below to knoiw sample queries, how the system interprets them, and what you can expect after applying them.
| Use case | Example query | System interpretation | What gets created |
|---|---|---|---|
| Create New active users | Users who installed the app in the last 7 days and opened the app at least once | App Install date within last 7 days AND App Open event occurred | A segment with install timeframe and recent engagement conditions. Reachable user count is shown before applying. |
| Create Cart abandonment | Users who added items to cart but did not complete purchase in the last 3 days | Add to Cart event occurred AND Purchase event NOT occurred within 3 days | A behavior‑based abandonment segment ready for recovery campaigns. |
| Create Inactive users | Users who have not opened the app in the last 30 days | App Open event NOT performed in the last 30 days | An inactivity‑based segment suitable for re‑engagement journeys. |
| Create High‑intent shoppers | Users who viewed a product more than 3 times in the last week | Product Viewed event count ≥ 3 within last 7 days | A high‑interest browsing segment for promotions or recommendations. |
| Create Recently converted users | Users who made a purchase in the last 14 days | Purchase event occurred within last 14 days | A recent buyers segment for cross‑sell or onboarding campaigns. |
Best practices for writing good queries
To get the best results:
- Use simple and clear language.
- Mention timeframes explicitly (last 7 days, last month, etc.)
- Clearly state did vs did not actions.
- Avoid combining too many conditions in one sentence.
- Prefer business intent over technical terms.
Good: “Users who purchased in the last 30 days but haven’t opened emails.”
Less effective: “Create a complex segment of users who might be inactive.”
Edge cases to be aware of
- Very broad requests may return extremely large audiences.
- Very complex sentences may result in partial interpretation.
- Unsupported attributes or events may be ignored.
- Conflicting conditions can reduce audience size unexpectedly.
- Always review the generated rules before saving.
Updated 2 months ago