1 / 8
or click to navigate
Adobe Summit 2026 FullStory × Adobe Session Intelligence

From Abandoned Session
to Personalized Campaign

How FullStory's Profile & Session Summary APIs generate the exact data payload Adobe needs to power hyper-personalized remarketing emails — automatically, at scale.

Anywhere: Activation Session Summary API Profile API Adobe Campaign
The Concept

One Session. One Payload. One Perfect Email.

Traditional remarketing sends everyone the same generic "you left something behind" email. FullStory reads what actually happened in the session and outputs structured data Adobe can use to personalize every variable.

🚗
User Session
CarGo browsing & abandonment
🧠
FullStory AI
Session Summary API + Profile
📦
Structured Payload
Deterministic JSON output
📧
Adobe Campaign
Personalized remarketing
Open Live AJO Workflow
Step 1 — The Session

Rachel browses CarGo, then disappears.

She searched for cars in Paris, clicked through an Audi S6 and two other vehicles, repeatedly changed dates, rage-clicked on the price — and left without booking. 313 events. No conversion.

What FullStory sees that Adobe can't: Rachel rage-clicked the Total Price element 10 times and the Pickup Location 9 times. She changed dates across three vehicles. That's not indifference — that's friction.
CarGo homepage

CarGo homepage — entry via Seasonal UTM Campaign

Vehicle detail page

Vehicle detail — browsed Audi S6, Toyota Avalon, Bentley

Step 2 — The Profile

Configure exactly what Adobe receives.

A FullStory Summary Profile is a reusable config that tells the AI what to extract and how to structure the output. One profile. Applied to every abandoned session. Same schema every time.

Why temperature: 0.1? Adobe's template engine needs the same field names and types every time. Low temperature pins the model to extraction over invention.
FullStory Create Profile API
Adobe Remarketing Profile POST /v2/visit_profile
// Profile ID: ca898f09-...
{
  "name": "Adobe Remarketing — Email Variables",
  "slice": { "mode": "LAST", "duration_limit_ms": 3600000 },
  "context": { "exclude": ["org", "location"] },
  "events": {
    "trim_to_last_n_selectors": 2,
    "include_descriptions": true
  },
  "llm": {
    "pre_prompt": "You are analyzing a car rental session...",
    "post_prompt": "Populate every field using only what
    you observed in the session...",
    "temperature": 0.1,
    "response_schema": { /* enforced */ }
  }
}
Step 3 — The Payload

One API call. Every variable Adobe needs.

Call GET /v2/sessions/{id}/summary and get deterministic JSON — ready to pipe into Adobe Campaign.

36 rage clicks. Frustration level: high. The API recommends an urgent tone with a 15% discount — all derived from session behavior, not rules engines.
Quick:
API Response — Adobe Remarketing v4
// GET /v2/sessions/{id}/summary?config_profile=f7ee...

{
  "response": {
    "product_page_1": {
      "vehicle": "Audi S6 2016",
      "time_on_page_seconds": 4
    },
    "product_page_2": {
      "vehicle": "Toyota Avalon",
      "time_on_page_seconds": 38  // → primary interest
    },
    "product_page_3": {
      "vehicle": "Bentley Trailers",
      "time_on_page_seconds": 0
    },
    "primary_interest_make":  "Toyota",
    "primary_interest_model": "Avalon",
    "primary_interest_year":  "2022",
    "primary_interest_price_per_day": "$130",
    "primary_interest_category":      "sedan",
    "abandonment_point": "vehicle_detail",
    "frustration_signal_counts": {
      "rage_clicks": 36,
      "dead_clicks": 2
    },
    "frustration_level":     "high",
    "recommended_email_tone": "urgent",
    "recommended_offer": "15% off — expires in 48h"
  }
}
Step 4 — The Email

Adobe reads the payload. The email writes itself.

Every highlighted value maps directly to a field in the FullStory payload. No manual segmentation. No guesswork.

frustration_level: high → empathetic tone, acknowledge friction first recommended_email_tone: urgent → lead with the apology, not the product time_on_page_seconds: 38 → "You spent 38 seconds" — proves we noticed recommended_offer → "15% off — expires in 48 hours" primary_interest_* → "Toyota Avalon 2022" woven into the save product_page_1 + _3 → "Audi S6 and Bentley" still available rage_clicks: 36 → triggered the "we owe you" subject line abandonment_point → vehicle_detail — never reached checkout
Step 5 — Delivered

Rachel opens her inbox. The email knows her.

Open AJO Workflow
📥 Inbox 9,095
⭐ Starred
🕐 Snoozed
➤ Sent
📄 Drafts 18
∨ More
Labels
📅 Calendar 5,191
✅ Closed/Won
📊 FullStory Digests
🔴 MRR Alerts 16
Primary
CarGo Rentals
Rachel, we owe you a better experience — here's 15% off the Avalon
2:14 PM
Airbnb
Your upcoming trip to Paris
11:02 AM
Chase Sapphire
Your March statement is ready
9:45 AM
Spotify
New music for your road trip playlist
Yesterday
▼ Every variable came from FullStory
Subject line
primary_interest_make + _model
$130/day price
primary_interest_price_per_day
38s dwell time
product_page_2.time_on_page_seconds
Discount code
recommended_offer
Also browsed
product_page_1 + _3 .vehicle
Urgent tone
recommended_email_tone
48hr expiry
rage_clicks: 36 → high frustration
Sedan category
primary_interest_category
The Takeaway

FullStory doesn't just record sessions.
It turns them into signals.

The same profile runs on every abandoned session. The output schema never changes. Adobe always gets exactly what it needs — and every email feels like it was written by a compassionate customer service agent.

POST /v2/visit_profile GET /v2/sessions/{'{id}'}/summary temperature: 0.1 response_schema: enforced
Try It Live — Session Summary Demo