GLP-1 Journey Personas

What this is

Built from observed behaviour

Most GLP-1 datasets are modelled from pharmacy claims or inferred from survey panels. This one is derived from what consented participants actually did across ChatGPT, Gemini, Chrome, Google Search and Amazon: the questions they asked, the pages they read, the things they bought. Those events are resolved to one profile per person, then rolled up into journey, barrier, brand and momentum blocks.

Why it is different

Four things a claims or survey cut cannot give you

What is in each profile

150 features across six blocks

Journey

journey_stage (current), journey_max_stage (furthest reached), journey_activity_status (active, lapsed, discontinued), journey_stage_path, journey_funnel_velocity, per-phase dwell, journey_reinitiation_count.

Brand and drug

brand_metrics_drug_primary, route (injectable or oral), brand breadth, and brand_metrics_ai_introduced_brand_flag, which marks a brand the AI raised that the person never named.

Barriers

barriers_cost, barriers_side_effect_fear, barriers_needle, barriers_access_supply and others. Each is a salience score between 0 and 1.

Trajectory

trajectory_engagement_slope, trajectory_at_risk_score, and month-over-month mom_* momentum, volatility and peak timing.

Information

information_private_public_gap, information_private_share, information_source_concentration.

Coverage

coverage_cohort_depth (aware, engaged, on-drug), per-surface coverage flags, and relevance_cf, a graded GLP-relevance confidence score.

How teams use it

Five starting points

Sample query

On-drug audience by demographics, barrier and drug

SELECT
    g."age_group",
    g."brand_metrics_drug_primary"                       AS primary_drug,
    COUNT(DISTINCT g."user_id")                          AS unique_personas,
    ROUND(AVG(g."barriers_cost"), 3)                     AS cost_salience,
    ROUND(AVG(g."barriers_side_effect_fear"), 3)         AS side_effect_salience,
    ROUND(AVG(g."trajectory_at_risk_score"), 3)          AS avg_at_risk
FROM MEASURE_EVENT_METRICS_PROD.ACCESS.GLP1_JOURNEY_PERSONAS g
WHERE g."coverage_cohort_depth" = 'on_drug'
GROUP BY g."age_group", g."brand_metrics_drug_primary"
HAVING COUNT(DISTINCT g."user_id") >= 20
ORDER BY unique_personas DESC;

The HAVING clause in this example limits results to groups of at least 20 people.

Specifications

What ships

Privacy & provenance