Salesforce Data Cloud Unified Customer Profile — Complete Guide 2026 | Module 07
The Unified Customer Profile
Complete Guide 2026
Master the single most important object in Salesforce Data Cloud — what the Unified Customer Profile contains, how it is used and how every Salesforce cloud benefits from it
- What Is the Unified Customer Profile?
- What the Unified Profile Contains — All Data Types
- A Real Unified Profile — Complete Example
- Individual vs Unified Individual — Critical Difference
- How the Unified Profile Is Used Across Salesforce
- The Profile API — Querying Profiles Programmatically
- Data Graphs — The Relationship Map on the Profile
- Profile Freshness — How Current Is the Data?
- Real-World Profile Use Cases
- Common Unified Profile Mistakes
- Quick Quiz
- Interview Questions for This Module
The Unified Customer Profile is the single, complete, deduplicated record of a customer that Data Cloud creates after Identity Resolution merges all matching records from different source systems. It is the most important object in Data Cloud — every downstream feature depends on it.
Before Data Cloud, your customer existed as fragments. A Contact in Salesforce CRM. A Subscriber in Marketing Cloud. A User in your mobile app. A record in your ERP. Each fragment had different information and none of them were connected. Your marketing team saw a different customer than your service team. Your sales rep had no idea what the customer complained about last week.
The Unified Customer Profile ends all of that. It is one record that combines the best information from every source — giving every team, every Salesforce cloud and every AI agent the same complete view of the same customer. When a customer calls support, the agent sees their entire purchase history. When Agentforce responds to a chat, it knows the customer's loyalty tier, open cases and recent behavior. When Marketing Cloud sends a campaign, it targets based on the customer's full cross-channel engagement.
The Unified Profile Is Like a Doctor's Comprehensive Patient File
Imagine visiting a hospital for the first time after years of seeing different doctors, specialists and pharmacies. Each provider has their own records — your GP has your basic health history, the cardiologist has your heart scans, the pharmacy has your medication list, the lab has your blood test results.
A great hospital creates a comprehensive patient file that pulls everything together — your complete medical history, all medications, all test results, all specialist notes — in one record any doctor can access instantly. When a new doctor opens your file they have your complete picture. They do not ask you to repeat your history. They do not prescribe something that conflicts with your existing medication.
That is the Unified Customer Profile. Every interaction your customer has ever had with your business — every purchase, every email, every support case, every website visit — combined into one complete record that any team or AI agent can access instantly.
INDIVIDUAL
Every team using this profile sees the same Priya. The service agent knows she is a Gold VIP with 47 orders and no current complaints. The marketing team knows she is highly engaged via email with Electronics affinity and an active cart. Agentforce knows her complete context before saying a single word. This is what makes Data Cloud transformational — one profile, every team, complete context.
| Factor | Individual DMO | Unified Individual DMO |
|---|---|---|
| What it is | Source record from one specific system | Merged master record created by Identity Resolution |
| How many per customer | Many — one per source system | One — per real-world customer |
| Data quality | Raw — reflects source system | Best values from all sources via Reconciliation Rules |
| Created by | DLO to DMO field mapping | Identity Resolution after matching and reconciliation |
| Used for Segmentation | No — segments use Unified Individual | Yes — all segments filter on Unified Individual |
| Used for Insights | Inputs via related DMOs | Calculated Insights stored directly on Unified Individual |
| Relationship | Source records linked TO Unified Individual | Contains merged attributes FROM all Individual records |
| Count Example | 10 million Individual records from 4 sources | 3 million Unified Individuals — 70% merge rate |
Interviewers frequently ask whether you can build a segment on the Individual DMO. The answer is NO. Segments in Data Cloud can only be built on the Unified Individual — not the raw Individual DMO. This is because the Unified Individual is the deduplicated, complete record. Building segments on raw Individual records would result in the same customer being counted multiple times and receiving duplicate communications.
What Is the Profile API?
The Data Cloud Profile API is a REST API that allows any external application to query a customer's Unified Profile in real-time. Instead of waiting for a batch export or activation to push data, the Profile API lets your website, mobile app, call center tool or Agentforce action pull a customer's complete profile the moment it is needed.
This enables true real-time personalization. When a customer opens your website the site queries the Profile API with their email or customer ID and receives back their complete unified profile in milliseconds — including their latest segment memberships, Calculated Insights and behavioral history. The website then renders personalized content based on that profile without any batch processing delay.
| Profile API Use Case | How It Works | What Is Returned |
|---|---|---|
| Website Personalization | Customer logs in → site queries Profile API with email → returns profile in real-time | LTV score, product affinity, segment memberships, last viewed items |
| Mobile App Context | App launch → queries Profile API with user ID → returns profile | Loyalty tier, points balance, active cart, personalized offers |
| Call Center CTI | Customer calls → phone matched to profile → agent sees full history | Complete interaction history, LTV, open cases, sentiment score |
| Agentforce Actions | Agent receives user query → Profile API query → uses profile for response | All unified profile data for RAG grounding |
| In-Store POS | Customer scans loyalty card → Profile API query → cashier sees full profile | LTV, purchase history, birthday offers, churn risk flag |
Profile API Key Technical Details
- Authentication via OAuth 2.0 using a Connected App
- Query by email address, phone number, Individual ID or Loyalty Card number
- Response includes all mapped DMO attributes and Calculated Insights
- Response time typically under 200ms for real-time use cases
- Rate limits apply — high-volume use cases require architectural consideration
- Data Space scoping — each query runs in the context of a specific Data Space
What Is a Data Graph?
A Data Graph is a pre-built semantic relationship map that defines how DMOs connect to each other around the Unified Customer Profile. Instead of Agentforce running 5 separate queries to retrieve an account's orders, cases, contacts, emails and products — a Data Graph pre-maps all those relationships so the complete context is retrieved in a single structured call.
Data Graphs are specifically designed to power Retrieval Augmented Generation (RAG) in Agentforce. When an AI agent needs to respond to a customer, it queries the Data Graph to get the complete customer context — all related data — in one efficient call. This context is then injected into the AI prompt so responses are grounded in real customer data rather than AI general knowledge.
Example Data Graph Structure
A typical retail Data Graph centered on Unified Individual might look like this:
Unified Individual (root) → Contact Points (email, phone, address) → Sales Orders (last 12 months) → Order Products (line items per order) → Web Cart (active cart items) → Email Engagement (last 30 days) → Service Cases (all time) → Calculated Insights (LTV, churn, affinity)
When Agentforce queries this Data Graph for a specific customer, it gets all of the above in one call. The agent now has complete context — what they bought, what is in their cart, their email history, any open cases and their value score — before saying a single word to the customer.
| Data Graph Property | Detail |
|---|---|
| Created In | Data Cloud Setup → Data Graphs → New |
| Root Object | Always starts from Unified Individual — the customer is the anchor |
| Relationships | Defined by connecting DMOs via shared key fields |
| Used By | Agentforce for RAG grounding, Profile API for related data retrieval |
| Performance | Pre-mapped structure means faster retrieval than ad-hoc queries |
| Depth | Can traverse multiple relationship levels — Order → Order Product → Product |
The Unified Customer Profile is only as current as the most recently processed data. Profile freshness depends on three factors — how often Data Streams ingest new data, how often Identity Resolution runs and how often Calculated Insights are refreshed. Each adds latency.
| Data Type | Ingestion Mode | Profile Update Latency | Impact |
|---|---|---|---|
| Website cart events | Streaming | Seconds | Abandoned cart triggers fire instantly |
| Mobile app events | Streaming | Seconds | In-session personalization works |
| CRM Contact updates | Batch — daily | Up to 24 hours | Address changes take a day to appear |
| Order history | Batch — daily | Up to 24 hours | Yesterday's orders on profile by morning |
| Email engagement | Batch — hourly | Up to 1 hour | Recent opens visible within the hour |
| LTV Calculated Insight | Scheduled compute | Hours — depends on schedule | LTV reflects orders from last compute cycle |
| Churn Score | Daily Einstein compute | 24 hours | Churn score is always from yesterday |
| Segment Membership | Full or Rapid refresh | 15 min (Rapid) to daily (Full) | Segment gates are not always real-time |
Design your ingestion strategy around the business requirement for each data type. Cart events MUST be streaming because a 1-hour-old cart event is useless for abandoned cart recovery. But a customer's home address does not need to be streaming — daily batch is perfectly sufficient and costs a fraction of the credits. The Unified Profile is always a composite of data at different freshness levels — the goal is to ensure each type of data is as fresh as the business use case requires it to be.
🛒 Retail — VIP Customer Recognition at Every Touchpoint
A premium fashion retailer used the Unified Profile to create consistent VIP treatment across every channel. When a Gold loyalty member walked into a store and scanned their card at the POS, the system queried the Profile API and showed the store associate their online purchase history, recent browse behavior and preferred brands. The associate could immediately say “I see you were looking at the new autumn collection online — we just received the pieces you viewed.” When the same customer called support, the agent saw they had spent over 2 lakh rupees in the last year and immediately escalated them to the dedicated VIP support queue. When they received a marketing email it was personalized to their exact preferred brands and sizes. One unified profile — consistent VIP experience everywhere.
🏭 B2B SaaS — Proactive Retention Before Churn
A SaaS company used the Unified Profile's Calculated Insights to identify accounts showing churn signals 90 days before renewal. The churn score was computed from product usage events (low login frequency, declining API calls), marketing engagement (unopened emails for 45+ days) and support history (multiple unresolved tickets). When a profile's churn score crossed 0.75, a Data Action fired automatically. It created a high-priority task for the Customer Success Manager, enrolled the account in a retention email sequence and flagged the opportunity in Salesforce Sales Cloud for executive review. The result was a 28% improvement in retention rate for accounts that would previously have churned without warning.
🤖 Agentforce — Personalized AI with Complete Context
An insurance company deployed Agentforce for customer service on their website chat. Without Data Cloud integration the agent gave generic responses — asking customers to repeat information already on file, unable to reference past interactions. After integrating with the Unified Customer Profile via Data Graph, the Agentforce agent greeted each customer by name, knew their current policy details, could see any open claims, understood their payment history and had access to their previous chat transcripts. When a customer asked “I want to update my address” the agent said “Of course Priya — I can see your current address is in Mumbai. What is your new address?” instead of “What is your customer ID and current address?”. Customer satisfaction scores for the AI channel increased by 41% within 3 months of the integration.
Mistake 1: Building segments on Individual DMO instead of Unified Individual
Attempting to create a segment that filters directly on the raw Individual DMO. This is not possible in Data Cloud — segments can only be built on Unified Individual. Teams that do not understand this distinction spend time troubleshooting why their segment criteria are not visible. Always build segments on Unified Individual — it is the only segmentation target for person-based audiences.
Mistake 2: Expecting the Unified Profile to always reflect the latest data
Assuming that a purchase made 10 minutes ago is immediately visible on the Unified Profile for segmentation and marketing. Batch-ingested data — like CRM orders synced daily — will not appear until the next batch run. Calculated Insights computed on a daily schedule reflect yesterday's data, not today's. Communicate data freshness expectations clearly to business stakeholders to avoid confusion when the profile does not show the most recent interaction.
Mistake 3: Not including Consent data in the profile strategy
Building a beautiful Unified Profile with rich behavioral and transactional data but forgetting to map the Contact Point Consent DMO. Activating segments to Marketing Cloud without consent tracking means opted-out customers receive marketing communications — violating GDPR, CAN-SPAM and other regulations. Consent management must be part of the profile design from day one — never retrofitted after go-live.
Mistake 4: Not configuring Data Graphs before implementing Agentforce
Deploying Agentforce without configuring Data Graphs in Data Cloud. Without Data Graphs, AI agents can only access basic profile attributes — they cannot retrieve related order history, open cases or behavioral context in one call. The result is an Agentforce agent that knows a customer's name but cannot access their purchase history — providing shallow, generic responses that frustrate customers. Data Graphs must be configured before Agentforce is given access to Data Cloud profile data.
Mistake 5: Confusing Unified Individual count with match rate success
Celebrating a high Unified Individual count as a sign of success. In fact a high Unified Individual count relative to raw Individual count means low merge rate — most records are NOT being matched. A successful implementation should have significantly fewer Unified Individuals than raw Individual records. If you have 10 million raw Individual records and 9.5 million Unified Individuals — Identity Resolution is only matching 5% of records. That is a problem to investigate, not a milestone to celebrate.
Q1: C — Unified Individual | Q2: B — Data Graph not configured | Q3: C — Up to 24 hours | Q4: B — Very low match rate | Q5: C — Data Cloud Profile API
The Unified Customer Profile is the single, complete, deduplicated record of a customer created after Identity Resolution merges all matching records from different source systems. A CRM Contact record contains only what was entered into Salesforce CRM — it has no awareness of the customer's Marketing Cloud email engagement, website browse behavior, mobile app interactions or ERP purchase history. The Unified Customer Profile combines all of these into one record — taking the best values from each source via Reconciliation Rules. It also includes Calculated Insights like LTV and churn score computed from cross-source data, segment memberships and consent data. Where a CRM Contact shows you who the customer is in Salesforce, the Unified Profile shows you who the customer is across your entire business ecosystem.
An Individual DMO record is a source record that came from one specific system — a CRM contact, a Marketing Cloud subscriber or an app user. The same real-world customer might have five Individual records across five different source systems. A Unified Individual DMO record is the merged master record created by Identity Resolution that combines all matching Individual records into one deduplicated profile using Reconciliation Rules to resolve field conflicts. You can have 10 million Individual records but only 3 million Unified Individuals — because Identity Resolution merged the duplicates. Segments in Data Cloud use exclusively the Unified Individual — never the raw Individual DMO. This is critical because segmenting on Individual would count the same customer multiple times if they exist in multiple source systems, resulting in duplicate communications and inflated audience counts.
Data Cloud powers Agentforce through three mechanisms working together. First, the Unified Customer Profile provides the AI agent with complete customer context before it generates any response. When a customer contacts support via chat, the agent queries the Data Graph — which is a pre-mapped relationship structure connecting the Unified Individual to all related DMOs like orders, cases, email engagement and calculated insights — and retrieves the full customer picture in one call. This is the RAG grounding mechanism — the AI response is generated based on real customer data rather than general AI knowledge. Second, Calculated Insights on the profile — like churn score and LTV — drive intelligent routing decisions. A high-value high-churn-risk customer is automatically escalated to a human agent rather than handled by the AI alone. Third, segment memberships on the profile tell the agent which audience the customer belongs to — allowing the agent to apply segment-specific offers, messaging and service levels.
Profile freshness is a composite — different data types have different latencies depending on ingestion mode and processing schedule. Streaming data like cart events and mobile app interactions appears on the profile within seconds. Hourly batch data like email engagement appears within an hour. Daily batch data like CRM Contact updates and order history appears within 24 hours. Calculated Insights computed on a daily schedule reflect yesterday's data. Segment membership updates on Rapid Refresh take 15 minutes and on Full Refresh can take hours. I design for freshness by matching the ingestion mode to the business requirement for each data type. Cart abandonment recovery requires streaming — a 1-hour-old cart event cannot trigger a relevant recovery message. But a customer's home address has no real-time requirement — daily batch is perfectly appropriate and costs a fraction of streaming. The key stakeholder communication is that the profile is always a composite of data at different freshness levels — not a perfect real-time mirror of the customer at this exact moment.
I would design a CTI integration that queries the Data Cloud Profile API the moment an inbound call is received. Using the caller's phone number as the lookup key, the Profile API returns the complete Unified Profile within milliseconds — before the agent even picks up the call. The agent's screen immediately shows the caller's full picture. Their name and account status. Their LTV and loyalty tier. Their last three purchases including amounts and products. Any open support cases with their current status. Their churn risk score if elevated. Their email engagement indicating recent interaction. From the Calculated Insights on the profile, the system automatically routes the call to a senior agent if the customer is VIP Gold tier with LTV over one lakh rupees. The Agentforce agent then greets them by name, references their specific situation and can immediately discuss their open cases without making the customer repeat information they have already provided. The customer experience transforms from “may I have your account number please” to “Good afternoon Priya, I can see your order from last week — is that what you are calling about?”