HARBINGER AI

Case Study

Texas Water Development Board

Public Opinion & Water Conservation Study

Methodology

At-a-glance process from survey architecture to activation-ready persona conversation.

Methodology Overview

This workflow shows how survey data becomes activation-ready personas you can interact with.

1

Survey Design

Questions are intentionally structured for both segmentation and driver modeling.

2

Segmentation

Clustering reveals distinct audience groups with different needs and behaviors.

3

Driver Modeling

Regression identifies which perceptions most influence the business outcome.

4

Persona Build

Each segment is translated into a clear persona and messaging direction.

5

Enrichment

Media and behavior context is layered in to make each persona activation-ready.

6

Persona Conversation

Teams can run simulated participant conversations for copy testing, feedback, and focus-group style interviews.

Survey Results

Start with the source responses. Fields are organized by purpose so teams can clearly separate segmentation inputs, driver-model inputs, profile/enrichment signals, and context fields.

Survey Input Design

Questions are tagged by intent so teams can quickly separate segmentation inputs from driver-model inputs.

Segmentation Inputs: - Driver Inputs: -

Segmentation Questions

Driver Questions

Source

Loading...

Total Rows

-

Filtered Rows

-

Total Columns

-

Visible Columns

-

Search all fields Filter field Filter value Question purpose Rows per page

Loading data...

Showing 0 rows

Page 1 / 1

Model Outputs

Explore segmentation structure and regression drivers for each audience group.

Workflow Stage

Model Output

segmentation + regression

Segmentation Inputs

-

- driver-model inputs

Profiles Scored

-

records in this case

Audience ID Field

-

- audience groups

Section A

Step 1: Segmentation Output

See the audience groups created from the segmentation design questions.

Segmentation Input Questions

Biggest environmental problem in Texas (Q1)
Who uses the most water (Q13)
First association with water in Texas (Q2)
Could respondent do more to conserve (Q16)
Water bill concern intensity (Q4)

Audience Group Mix

Audience Group	Audience Share	Respondents

Audience Group Distribution

Distribution from the audience assignment field.

Audience Feature Heatmap

Side-by-side view of key segmentation attributes by audience group.

Audience Positioning Map

Interactive attribute map built from segmentation questions only.

Built from segment-separating attribute signals.

Active Audience Structure

Hover a bubble to see how strongly it aligns to each audience group.

Section B

Step 2: Regression Drivers by Audience Group

Compare all positive and negative drivers across every audience group.

Interpretation Notes

Use this tab to understand group structure and modeled drivers, then move to Segment Playbooks for persona strategy and activation guidance.

Segment Playbooks

Persona cards with messaging, motivators, barriers, and media behavior guidance for each audience group.

Loading audience groups...

Source

Loading...

Total Rows

-

Audience Groups

-

Audience ID Field

-

Persona Card Gallery

Select any persona card to open the full strategy profile.

Select an audience group

-

Respondents

-

Audience Share

-

Avg Q4

-

Avg Q5

-

Mindset

-

Messaging Strategy

-

Motivators

-

Barriers

-

Brand Opportunity

-

Media Behavior

-

Activation Direction

-

Signature Survey Signals

Q1: -

Q13: -

Q16: -

Q17: -

Audience Matrix

Side-by-side comparison across audience groups.

Audience Group	Share	Top Signal 1	Top Signal 2

Audience Activation

Project reachable audience size and prepare activation-ready output.

Waiting for audience data...

Audience Group Total Addressable Audience Match Rate (0 to 1) CPM ($)

Destination DSP

Choose where this audience package should be sent.

The Trade Desk

Google DV360

Amazon DSP

Xandr

Projected Matched Audience

-

Planned Impressions (x3)

-

Estimated Spend

-

Audience Reach Plan

Audience Group	Audience Share	Projected Matched

DSP Payload

Demo output ready to post to your activation endpoint.

Public Opinion & Water Conservation Study

Methodology

Methodology Overview

Survey Design

Segmentation

Driver Modeling

Persona Build

Enrichment

Persona Conversation

Survey Results

Survey Input Design

Segmentation Questions

Driver Questions

Loading...

-

-

-

-

Model Outputs

Model Output

-

-

-

Step 1: Segmentation Output

Segmentation Input Questions

Audience Group Mix

Audience Group Distribution

Audience Feature Heatmap

Audience Positioning Map

Active Audience Structure

Step 2: Regression Drivers by Audience Group

Regression Driver Output by Audience Group

Segment Playbooks

Loading...

-

-

-

Persona Card Gallery

-

-

-

-

-

Mindset

Messaging Strategy

Motivators

Barriers

Brand Opportunity

Media Behavior

Activation Direction

Signature Survey Signals

Audience Matrix

Creative Outputs

Generated Creative Gallery

Generation Notes

Audience Activation

Destination DSP

-

-

-

Audience Reach Plan

DSP Payload

Activation Framework

Customer Journey

Live Communications

Communication Lanes

Connector Workspace

Journey Copilot

Persona Lab