Adobe Summit 2026
Session Guide

Content Compass · AI Agent Forum Working Group · Three participation tracks
Live session Choose your track Bring your agent
Access the prototype
Track 1 · Prototype
Content Compass Agent Prototype
Use Adobe's Content Compass Agent prototype to create personalised research journeys across Business.Adobe.com. Explore how the intelligence layer — personas, themes, and product intelligence rubrics — shapes the content recommended to you based on your priorities.
What you'll do
1
Open the prototype
Launch Content Compass in your browser

Open the Content Compass prototype link shared in the session. No login required — it runs entirely in your browser. If you want live AI-powered matching, paste your Anthropic API key in the Connect bar at the top.

2
Enter your priorities
Type in your real priorities — or use a sample persona

Fill in one, two, or all three prompt fields. The more specific you are, the better the journey. Or choose a sample persona from the dropdown to see an example output instantly.

💡 Tip
Be specific about your role, current tech stack, and what problem you're trying to solve. "CMO at a B2B SaaS company trying to prove marketing ROI" will produce a much more tailored journey than "marketing leader."
3
Explore your journey
Review, refine, and compare
  • Read the Interpretation Box — see how the system read your priorities
  • Try the Explore Themes panel to add adjacent topics
  • Switch between Timeline and Detail views
  • Use ⇄ Compare Journeys in the header to see how two different personas or priorities produce completely different journeys
4
Share back
Submit feedback on your journey

At the bottom of your journey card, use the Share feedback box to tell us: was this relevant? What was missing? What surprised you? Your input shapes what we build next.

Sample inputs to try
CMO · B2B Tech
Prove ROI + scale with AI
Priorities
Prove marketing ROI to the board, accelerate pipeline, build personalisation at scale without adding headcount.
Capabilities
Unified customer data, AI journey orchestration, content that converts.
Expect: RTCDP + AJO + CJA plays, Initial Research arc
Marketing Leader · B2B SaaS
Demand gen + AI content
Priorities
Scale demand gen, improve lead quality, use AI to reduce manual campaign work.
Capabilities
Better segmentation, automated nurture, AI content at scale with brand guardrails.
Expect: Marketo + GenStudio + AJO plays
Discussion prompts
1
Did the journey match your actual priorities? Where did it over- or under-index?
2
If an AI agent built this journey on your behalf — without you typing anything — what would it need to know about you?
3
What would make this useful enough to actually follow as a research plan?
Track 2 · Full
Build a Content Compass Skill — Full Version
Navigate to the Content Compass Lab in the Adobe Playground. Your agent can access files containing the full Content Compass intelligence — rubrics, personas, themes, and scored content. Replicate the process of building an Adobe research skill with your own agent using the complete dataset.
⚡ Before you start
  • Have your AI agent open and ready (Claude.ai, ChatGPT, Copilot, or your team's agent)
  • Navigate to the Content Compass room in the Content Compass Lab
  • Your agent will need file access permissions to read the 5 intelligence files
How it works
1
Access the Lab
Go to the Content Compass Lab

The Content Compass Lab contains the full Content Compass intelligence files — persona rubrics, product intelligence rubrics, and the scored content library. Your agent can read these files directly.

2
Load the intelligence
Point your agent at the Content Compass files

Ask your agent to read the files and confirm it understands the structure — 9 personas, 25 product intelligence rubrics, 158 themes, and 906 scored content records — plus theme detection data for 1,375 items. This becomes your agent's "research skill."

Suggested prompt to your agent
"Read the Content Compass intelligence files in this room. Understand the persona rubrics, product intelligence rubrics, and scoring logic. Then I'll give you a buyer profile and ask you to build a research journey."
3
Build a journey
Give your agent a buyer profile and ask for a research journey

Use your own priorities or one of the sample profiles below. Ask your agent to: match a persona, identify relevant product rubrics, detect your journey stage, and recommend 4 weeks of content.

Suggested prompt
"I'm a VP Marketing at a B2B tech company. My priorities are proving marketing ROI, scaling content with AI, and building personalisation without adding headcount. Using the Content Compass rubrics, build me a 4-week research journey."
4
Refine and compare
Ask your agent to explain its reasoning and iterate
  • Ask it to explain which persona it matched and why
  • Ask it to try a different persona or play and compare the output
  • Ask it what content it would prioritise if you were in a later evaluation stage
  • Try: "What themes from the rubric are most relevant to my priorities?"
5
Reflect
Compare your agent's output to the Content Compass prototype

Open the Content Compass prototype (Track 1) and enter the same priorities. How similar are the journeys? Where do they diverge? What did your agent do better or worse? What would improve it?

Discussion prompts
1
How did your agent's journey compare to the prototype? What was different — and why?
2
What would it take to turn this into a reusable skill your agent could use every time someone needs a research journey?
3
What intelligence was missing from the files that would have made your agent's recommendations more useful?
Files in the Content Compass Lab
Agent journey instructions — copy & use
For agents with file access

If your agent can access the Content Compass Lab files directly, you don't need this prompt — point it at the files instead. Use this prompt if your agent can't access files or you want a self-contained version with all rubric logic embedded.

Full rubric prompt embedded — click "Copy prompt" to copy all 9 personas, 25 product intelligence rubrics, scoring logic, and journey arc instructions to your clipboard.
Track 3 · Lite
Build a Content Compass Skill — Lite Version
Navigate to the Content Compass Lab in the Adobe Playground where it can access a slimmed-down text excerpt of Content Compass Intelligence. Replicate the process of building an Adobe research skill for a focused set of content formats and Adobe offerings — no file access needed, just paste and go.
✓ What's included in the Lite version
  • 3 personas: CMO, Marketing Leader, MarTech Leader
  • 6 product offerings: GenStudio, RTCDP, CJA, AJO, Workfront, Overall AI
  • Content formats: Reports, Articles, Customer Success Stories
  • Works with any agent — just paste the prompt text below
How it works
1
Navigate to the Content Compass Lab in the Adobe Playground
Find the Content Compass Lab

The Lite version uses a plain text excerpt — no file download needed. You can copy the intelligence prompt directly from the room, or use the prompt below.

2
Paste the intelligence prompt
Start a new conversation with your agent and paste the prompt

Open your agent (Claude.ai, ChatGPT, or any agent you have access to). Start a fresh conversation. Copy the full Lite Rubric Prompt below and paste it in. Your agent will read the rubrics and be ready to build journeys.

3
Build a journey
Replace the placeholder with your own priorities

At the bottom of the prompt you'll see [PASTE PERSON'S INPUT HERE]. Replace that with your real priorities, role, and context. Your agent will build a personalised journey using the rubric intelligence.

Try this input
"I'm a CMO at an enterprise B2C brand. I need to scale content production with AI, ensure brand consistency across global markets, and prove the ROI of our content investment."
4
Iterate
Ask your agent to refine and explain
  • "Focus more on customer success stories"
  • "I'm already evaluating vendors — shift to later-stage content"
  • "What themes from the rubric best match my priorities?"
  • "Build the same journey for a CJA-focused buyer and compare"
5
Reflect
What worked? What would you add?

Consider: if this lite rubric was a persistent skill in your agent — something it could call any time someone needed Adobe content guidance — how would you improve it? What's missing?

Lite rubric prompt — paste into your agent
How to use

Paste this entire prompt into a new conversation with your agent. It contains the scoring rubric for 3 personas, 6 product offerings, and the journey-building logic. No files needed — works in any agent with a reasonable context window.

CONTENT COMPASS LITE — ADOBE RESEARCH JOURNEY PROMPT

You are an Adobe content intelligence agent. Your role is to build a
personalised research journey using the rubrics below.

═══════════════════════════════════════════════════════
PERSONAS (3)
═══════════════════════════════════════════════════════

CMO (Chief Marketing Officer)
Key themes: Make/enable data-driven decisions; Maximize marketing impact/ROI;
Improve performance with real-time analytics; Eliminate data siloes;
Measure & improve customer journey; Drive creative excellence and brand integrity;
Create content more efficiently at scale; Prove marketing ROI;
Scale content creation with Gen-AI; Use/implement AI solutions that are business-safe

MARKETING LEADER (DM Marketing)
Key themes: Make/enable data-driven decisions; Maximize marketing impact/ROI;
Streamline data environments; Build and execute marketing programs;
Manage customer engagement across lifecycle; Measure & improve channel performance;
Scale content creation with Gen-AI; Leverage AI to optimize journey engagement;
Improve lead quality and conversion; Drive pipeline and revenue growth

MARTECH LEADER (MOps DM)
Key themes: Make/enable data-driven decisions; Streamline data environments;
Consolidate/streamline marketing technology stack; Improve operational efficiency;
Orchestrate work across tools and systems; Automate manual tasks with AI;
Ensure brand consistency; Maximize ROI of content; Streamline content workflows;
Use/implement AI solutions that are enterprise-ready

═══════════════════════════════════════════════════════
PRODUCT INTELLIGENCE RUBRICS (6)
═══════════════════════════════════════════════════════

GenStudio for Performance Marketing
Themes: Scale content creation with Gen-AI; Ensure brand consistency;
Drive adoption of Gen AI with risk reduction; Automate creation of content derivatives;
Streamline content/creative workflows; Maximize ROI of content;
Free up creative resources for strategic work; Get into market faster;
Use/implement AI solutions that are business-safe; Drive revenue growth

Adobe Workfront
Themes: Improve operational efficiency; Establish a marketing system of record;
Use/implement AI solutions that are business-safe; Maximize ROI of content;
Streamline content/creative workflows; Automate manual tasks with AI;
Orchestrate work across tools and systems; Reduce marketing dependency on IT;
Establish a unified view of marketing work; Increase marketing agility

Adobe Real-Time CDP (RTCDP)
Themes: Manage customer engagement across lifecycle; Collect and govern customer data;
Establish a unified view of customers; Turn customer data into real-time profiles;
Use customer data to drive personalized experiences; Modernize IT infrastructure;
Measure & improve marketing channel performance; Securely connect and analyze data;
Manage multiple sources of data easily; Enable real-time audience activation

Customer Journey Analytics (CJA)
Themes: Measure & improve marketing channel performance; Maximize ROI of campaigns;
Improve performance with real-time analytics; Uncover audience and journey insights;
Streamline/consolidate content performance data; Prove marketing ROI;
Assess strategic alignment of marketing work; Modernize data measurement;
Measure & improve content performance; Enable fast strategic shifts

Adobe Journey Optimizer (AJO)
Themes: Leverage AI to optimize journey engagement; Manage customer engagement
across lifecycle; Measure & improve marketing channel performance;
Build and execute marketing programs; Drive engagement through omni-channel
personalization; Enable real-time personalization; Automate personalized journeys

Overall AI (cross-product)
Themes: Automate manual tasks with AI; Leverage AI to improve organizational efficiency;
Leverage AI to optimize journey engagement; Scale content creation with Gen-AI;
Use/implement AI solutions that are business-safe & enterprise-ready;
Drive adoption of Gen AI; Automate personalized account journeys;
Improve decision-making with AI insights; Enable intelligent content recommendations

═══════════════════════════════════════════════════════
CONTENT TIERS
═══════════════════════════════════════════════════════

Initial Research (IR): Reports, blogs, thought leadership
→ Awareness, problem framing, building the business case

Go Deeper (GD): Articles, customer success stories, industry content
→ Proof points, real-world outcomes, deeper exploration

Get Specific (GS): Product pages, solution briefs, demos
→ Evaluation, feature comparison, decision support

═══════════════════════════════════════════════════════
HOW TO BUILD A JOURNEY
═══════════════════════════════════════════════════════

Step 1 — Match persona: Which of the 3 personas best fits this person's role?
Step 2 — Match products: Which 2-3 product rubrics match their priorities?
Step 3 — Detect stage: Are they early (exploring), mid (comparing), or late (evaluating)?
Step 4 — Build 4 weeks:
  Week 1: Initial Research — foundation building
  Week 2: Go Deeper — proof points and customer stories
  Weeks 3-4: Get Specific — evaluation and decision support
  (If late stage detected, start at Get Specific)

Step 5 — Return:
  1. Persona matched + reasoning
  2. Matched products (2-3) + why
  3. Journey stage
  4. 4-week plan with content types and topics per week
  5. What you'll learn: 3-4 outcome statements

═══════════════════════════════════════════════════════
NOW BUILD THE JOURNEY
═══════════════════════════════════════════════════════

[PASTE YOUR PRIORITIES, ROLE, AND CONTEXT HERE]
Resources in the Content Compass Lab
Content scores — Best Fit & Good Match
How to read this table

BF = Best Fit — top-tier signal, strongly aligned to this persona or product offering.
GM = Good Match — relevant but less specifically targeted. Share with your agent as context for scoring.

Clean BF list for agents

426 Best Fit items grouped by tier — formatted for pasting directly into an agent prompt or context window. No table formatting, just clean title + BF signal + URL.

906 items
Content CMO Mktg MktgOps GenStudio Workfront RTCDP CJA AJO AI
Discussion prompts
1
How did your agent do with just the rubric text? What did it get right vs. where did it struggle?
2
If you could add one more thing to the Lite prompt to improve the output, what would it be?
3
Would you use a prompt like this as a persistent skill in your agent? What would need to change?