Unified Co-Marketer

One agent that already knows your campaigns, journeys, and segments, ask in plain language and act on the answer.

Overview

The Unified Co-marketer is a single conversational agent within Netcore CE that brings together three previously separate agents:

  • Campaign Insight Agent
  • Journey Insight Agent
  • Segment Agent

The Unified Co-marketer provides a single conversational experience for campaign, journey, and segment analysis. Related questions can be asked within the same conversation, with context retained across follow-up questions.

The agent analyses campaign, journey, and segment data, explains performance and contributing factors, and provides recommendations based on the available data. Context is retained throughout the conversation, allowing follow-up questions to build on previous responses.

📘

How to Activate Co-marketer

Contact the Netcore account manager to enable the Unified Co-marketer.

Once enabled:

  • Ask Co-marketer appears in the top navigation bar of the Netcore CE dashboard.
  • Get Insights continues to appear on the Campaign and Journey listing pages.
  • The Co-marketer panel remains available on the Create Segment page.
  • No additional setup is required.

Changed Workflow

The capabilities previously available through the individual agents are now available through the Unified Co-marketer.

Earlier experienceWhere it lives now
Campaign Insight AgentCampaign performance, root-cause, content, STO, and audience-level analysis are available within the unified conversation.
Journey Insight AgentPortfolio ranking, journey comparison, drop-off, and node-level analysis are available within the unified conversation.
Segment AgentNatural-language segment creation is available from the Create Segment page and within the unified conversation.
Switching between agentsNo longer required. A single conversation can span campaign, journey, and segment data.

📘### What the Unified Co-marketer can read

The Unified Co-marketer works with campaign, journey, and segment data available within the current panel. It does not read data from CRM, order management, core banking, policy administration, or other external systems.

If a question depends on data outside these three domains, the agent indicates that the required data is unavailable rather than estimating an answer.

Key Capabilities

CapabilityDescription
Campaign Performance AnalysisAnalyse campaign performance across engagement, conversion, and revenue metrics. Identify top and bottom performers and compare performance across channels.
Root-Cause AnalysisAnalyse contributing factors such as audience composition, channel effectiveness, content, send timing, message frequency, fatigue, and device trends.
Audience InsightsAnalyse how different segments respond, identify audiences driving revenue, identify high-intent users missed during targeting, and detect audience overlap that may contribute to fatigue.
Journey Portfolio AnalysisAnalyse multiple journeys together, rank journeys by CTR, conversion, or revenue, identify the steepest drop-off, compare journeys with similar goals, and identify lifecycle stages that are not covered by any journey.
Journey and Node-Level AnalysisAnalyse progression within an individual journey, including stage-level performance, cadence gaps, dead branches, and nodes where users drop off.
Segment CreationDescribe an audience in natural language and convert the description into structured segmentation rules, with an estimated reachable-user count before saving.
Cross-Surface ReasoningConnect campaign, journey, and segment data within a single question without requiring manual data exports or report reconciliation.
Trend DetectionIdentify meaningful trends, anomalies, and patterns across campaign and journey performance, including unusual engagement spikes and declining metrics.
Content InsightsAnalyse the impact of creative elements such as subject lines, copy, tone, emojis, images, and calls to action on opens, clicks, and conversions.
Send-Time Optimization (STO) AnalysisEvaluate whether campaigns using Send-Time Optimization improved engagement compared with standard delivery.
Recommendations & Optimization OpportunitiesProvide recommendations for audience refinement, channel mix, frequency, timing, and content experimentation based on available performance data.
Brand-Aware InsightsWhen a brand profile is configured, analysis and recommendations reflect configured priority metrics, key segments, business terminology, and messaging guidelines.
Personalization & MemoryRetain preferences across conversations, including preferred metrics, channels of interest, reporting style, and answer format. These preferences are applied privately within the associated account.
Visual & Comprehensive ReportingPresent insights through summaries, tables, and charts, and consolidate portfolio-wide performance into a single business review when requested.

Access the Unified Co-marketer

The Unified Co-marketer can be accessed from the top navigation bar or from supported surfaces within Netcore CE.

From the top navigation

  1. Open the Netcore CE dashboard.
  2. Click Ask Co-marketer in the top navigation bar. The Co-marketer opens as a side panel over the current screen.
  3. To access the full experience, including agents and reports, click the Expand icon at the top right of the panel.
  4. Enter a question or select a suggested prompt.

From a listing page

  1. Navigate to Engage > Campaigns or Engage > Journey.
  2. Click Get Insights in the top-right corner of the listing page.
  3. The Co-marketer panel opens on the right and is scoped to the campaigns or journeys available on that page.

From the Create Segment page

  1. Navigate to Audience > Segments and click CREATE.
  2. The Co-marketer panel appears by default on the right side of the Create New Segment page.
  3. Describe the required audience or select one of the suggested prompts.

Sample Queries Asked

The following examples are organised by use case. Campaign names, journey names, channels, and timeframes can be adapted to the available data.

Campaign Performance

PromptWhat it provides
"What were the top 3 best and worst performing campaigns in the last 30 days?"A ranked view of the highest- and lowest-performing campaigns, along with the metrics supporting the ranking.
"Can you do a month-on-month analysis of the best-performing campaigns?"A period-over-period comparison showing the direction and size of performance changes.
"Can you help me do an in-depth analysis of the subject lines?"Subject-line patterns correlated with open and click performance for Email campaigns.
"Point out the best-performing STO campaign and compare all STO campaigns from the last 30 days."A comparison of Send-Time Optimization campaigns against each other and against standard delivery.
"Why did the New Year Mega Sale campaign underperform?"A root-cause analysis across audience, content, timing, frequency, and channel factors.

Audience and Segments

PromptWhat it provides
"Which segments are engaging the most, and which are dropping off?"Segment-level engagement and the audiences driving or diluting results.
"How did VIP customers respond compared with new users?"A side-by-side cohort comparison across engagement, conversion, and revenue.
"Show me add-to-cart users who were not targeted."High-intent users excluded from targeting, an estimated revenue opportunity, and options to Create Segment or Download List.
"Which segments are over-messaged?"Campaign overlap across segments and signals of potential audience fatigue.
"Create a segment of users who added to cart but didn't purchase in the last week."A structured segment definition with an estimated reachable-user count.

Journeys

PromptWhat it provides
"Generate an executive report for all journeys running in the past 180 days."A structured summary covering portfolio performance, top and bottom performers, and key trends.
"Give me my top 10 journeys by CTR in the past 180 days."A ranked list presented as a table and bar chart.
"Which journey has the highest drop-off rate?"The journey with the steepest drop-off, along with the node where the drop-off occurs.
"Compare all my cart abandonment journeys in terms of performance."A comparison across entries, conversion rate, and revenue for journeys with similar goals.
"Show me the channel mix across all my trigger-based journeys."A breakdown of channels used and their relative share of sends.

Across Campaigns, Journeys, and Segments

The unified conversation enables analysis across campaign, journey, and segment data.

PromptWhat it provides
"Which of my high-value segments are not covered by any active journey?"Cohorts that are not entering any journey, highlighting potential coverage gaps.
"My repeat buyers segment converted poorly last month. Was it the campaigns or the journeys?"A combined analysis across campaigns and journeys, with attribution where the available data supports it.
"Which segment received the most messages last week across both campaigns and journeys?"Total message volume per segment across campaigns and journeys, supporting frequency and fatigue analysis.
"Build me a segment of the cohort driving the highest revenue in my welcome journey."An audience derived from journey performance and created as a reusable segment.

Create a Segment Conversation

Segment creation is available within the unified conversation.

  1. Describe the required audience in natural language. For example: "Users who installed the app in the last 7 days and added items to cart but didn't purchase."
  2. The agent returns a card summarising the interpreted audience and an estimated Reachable Users count.
  3. Click View on the summary card to populate the segment builder. All conditions appear in the standard dropdown-based segmentation UI.
  4. Refine the segment by describing the required change in the same conversation. For example: "Change the timeframe to the last 14 days" or "Exclude users who made a purchase yesterday."
  5. After validating the rules, click Save.

The segment is then available like any other segment and can be used across campaigns and journeys.

👍### Important Points to Remember

  • Analysis is based on campaign, journey, and segment data available in the current panel.
  • Analysis is supported across Email, SMS, App Push, Web Push, WhatsApp, and RCS.
  • Only included segment IDs and list IDs are considered. Excluded IDs are not used in analysis.
  • For Email, the agent analyses subject lines. Subject lines are not analysed for WhatsApp, RCS, or SMS.
  • RCS message content is not analysed. A CTA recorded as zero indicates that no CTA was detected. Multiple CTAs are not tracked.
  • Key metrics tracked for campaign performance include Open Rate, Click Rate, Conversions, Revenue, and Unsubscribe Rate.
  • A broad prompt such as "Give me campaign data for the last 3 months" returns the first 5 campaigns by default. To retrieve more campaigns, specify the required number, for example "Show insights for the top 30 campaigns."
  • Segment creation is not available while editing an existing segment. The system generates one complete segment per submission.
  • Journey and campaign names displayed in tables and text responses are plain text. Clicking a name does not navigate to the corresponding journey or campaign.
  • Netcore recommends: Specify the number of campaigns or journeys required in the initial prompt for more focused results.

How the Unified Co-marketer works

The Unified Co-marketer interprets the request, retrieves the relevant data, identifies patterns, and presents the results in an actionable format.

  1. Understand the request
    Identifies the relevant metrics, campaigns, journeys, audiences, and time periods, along with the data domains involved.

  2. Analyse relevant data
    Evaluates performance, engagement, audience behaviour, journey structure, and historical trends. Where configured, the analysis also considers the brand profile and saved preferences.

  3. Generate insights
    Identifies relevant patterns, anomalies, relationships, and performance drivers from the available data.

  4. Deliver actionable answers
    Presents results in plain language with supporting metrics, observations, visualisations, and recommendations where applicable.

To explain performance, the agent considers multiple signals across three primary areas:

  • Content and tone: Content category, message tone, use of emojis, punctuation, alphanumeric characters, subject-line length, character count, and word count.
  • Timing: Day of the week, hour of the day, and day type, such as weekday or weekend.
  • Targeting: Whether the message is personalised and the campaign type.

Content Categories

The agent analyses message content and assigns campaigns to content categories to support like-for-like comparisons.

  • Promotion of New Arrivals: Keywords such as New Arrival, Fresh, Introducing, Breaking, and First Look.
  • Promotion of Recommended Products: Keywords such as Top Picks, Recommended, Must-Have, Trending, and Just For You.
  • Discounted Promotions: Keywords such as Sale, Discount, Flash Sale, Exclusive Offer, Price Drop, and Special Price.
  • Content Engagement: Keywords such as Guide, Tips, How To, Discover, Learn, and Insights.

Response Structure

The Unified Co-marketer can present results in several formats depending on the request.

Text summary:
A plain-language explanation of the findings, including notable patterns or anomalies.

Table:
For comparisons and rankings, results appear in a table with the relevant entity name, key metrics, and period.

Chart:
A visual representation of the data, typically a bar chart for rankings and a side-by-side view for comparisons.

Recommendation:
Where applicable, suggested next steps based on the analysed data.

⭐### Best practices for writing good prompts

  • Use simple and clear language.
  • Mention timeframes explicitly, such as last 7 days or last month.
  • Specify the number of results required, such as top 10 or top 30.
  • Clearly distinguish between actions users did and did not perform when describing an audience.
  • Avoid combining too many conditions in a single sentence.
  • Prefer business intent over technical terminology.
  • State the desired outcome or decision context explicitly.

Example:
"Compare engagement and revenue for VIP Repeat Buyers vs First-Time Buyers over the last 30 days."

Fallback Behaviour

When the requested data is unavailable or the intent is unclear, the Co-marketer responds with a clarifying message rather than estimating an answer.

For example, if a request refers to active users without defining what qualifies as active, the agent asks for clarification before proceeding.

Known Limitations

The following limitations apply to Unified Co-marketer insights:

  • For recently launched campaigns, insights are available after 24 hours.
  • Engagement and conversion data can continue to update for up to 30 days after a campaign. On the 30th day, the data becomes static. Recent campaigns may show lower CTR and Open Rate, which can affect insights.
  • Customer data uses a rolling one-year window. For example, if the current date is June 26, 2026, insights use data from June 26, 2025, through June 25, 2026.
  • Panel data refreshes on a scheduled cycle rather than in real time.
  • Very broad requests may return extremely large audiences.
  • Complex prompts may result in partial interpretation.
  • Unsupported attributes or events are ignored.
  • Generated segment rules should be reviewed before saving.

Responsible AI and Transparency

The Unified Co-marketer provides insights based on available campaign, journey, and segment data. Supporting metrics are presented alongside recommendations where applicable.

Generated insights should be reviewed in the context of organisational goals and business requirements before strategic decisions are made.

Data privacy and isolation: Each customer's data, brand profile, and saved preferences are isolated to the associated account and used only to answer queries for that account.

For more details, refer to AI Security and Governance.


Did this page help you?