Connecting AI assistants directly to your CRM unlocks seamless data interactions and automated workflows. By configuring Salesforce as a Model Context Protocol (MCP) server, you allow Claude AI to securely query data, perform actions, and interact with your Salesforce CRM in real time.

Here is a step-by-step guide to setting up a secure OAuth connection between Salesforce and Claude AI.

Step-by-Step Configuration

1. Create an External Client App in Salesforce 

First, set up an app in Salesforce to handle authentication.

• Go to Settings > Setup and search for External Client Apps.

• Click New External Client App, enter a name, and provide your email address.

• Scroll down to the API section and enableOAuth.

• Enter your Callback URL.

• Under Available OAuth Scopes, select:

  • Perform requests at any time

  • Access Salesforce-hosted MCP servers

• In the Security section, enable PKCE (Proof Key for Code Exchange) and Issue JSON Web Token-based access tokens for named users.

• Click Create.

2. Activate MCP Servers in Salesforce

• Search for MCP Servers under the API Catalog in Setup.

• Open the Subject All option and activate it.

• Copy the generated Server URL.

3. Configure Claude AI Connectors

• Log in to the Claude AI website, navigate to Customize, and open Connectors.

• Click Add > Add Custom Connector and give your connector a name.

• Paste the copied Salesforce Server URL into the Remote MCP Server URL field.

4. Authenticate and Connect

• Go back to your Salesforce External Client App settings.

• Copy the Consumer Key and paste it into Claude’s OAuth Client ID field.

• Copy the Consumer Secret and paste it into Claude’s OAuth Client Secret field.

• Click Add, then click Connect.

• When prompted by Salesforce, click Allow to authorize the connection.

Once authorized, you will see a "Connected to Salesforce" confirmation message. Claude AI is now ready to securely query records and execute CRM actions using the Model Context Protocol.

Traditional Salesforce is a lot like driving a standard car. It is undeniably powerful and reliable, but it requires a human in the driver's seat steering every turn, shifting every gear, and doing all the work manually.

To complete standard tasks in the traditional Salesforce interface, you spend a surprising amount of time navigating the platform rather than solving core business problems. Consider the everyday task of closing a service case:

• Step 1: Click the App Launcher icon.

• Step 2: Search for and select "Service."

• Step 3: Navigate to the Cases tab and locate the target record.

• Step 4: Click the "Close Case" action.

• Step 5: Manually update the status field to "Closed."

• Step 6: Select a resolution reason and click "Save."

While manageable, those six repetitive steps consume valuable minutes when multiplied across hundreds of open tickets every single day.

Enter Headless 360

Headless 360 completely eliminates the need to manually click through screens. Instead of a human operator working through the traditional user interface (UI), autonomous AI agents interact directly with Salesforce using background APIs.

The workflow shifts from complex UI navigation to a simple prompt:

1. Open your browser and navigate to an AI interface (such as Claude AI).

2. Type a single command: "Close case [ID] with reason: issue resolved."

That is literally it. The AI agent processes the natural language request, authenticates with Salesforce behind the scenes, makes the logical decision, and updates the record instantly. The entire process finishes in seconds, leaving the case fully closed without a single button click, page load, or manual data entry field.

This transition marks a massive leap in enterprise efficiency. By removing the visual interface overhead, Headless 360 transforms Salesforce from a tool you manually operate into an automated engine where AI agents execute back-office operations directly.

How to Set Up Essential Salesforce Company Settings

Ever wonder how global giants like Amazon, Nike, or Coca-Cola support millions of customers without falling into operational chaos? The secret lies in a well-configured Salesforce CRM.

Before building complex automations or building custom dashboards, every Salesforce administrator must configure the foundational settings of their Salesforce environment. Here is a guide to setting up your organization's digital headquarters.

Essential Company Settings to Configure

• Multi-Currency: 

  • If your business operates across different countries, enabling multi-currency allows sales teams to work in local currencies while Salesforce automatically converts values for reporting. Note: Enabling multi-currency is a permanent, non-reversible action.
    

  • Setup: Go to Setup > Company Information > Edit > check Activate Multiple Currencies > Save.

• Corporate Currency: 

  • This serves as your company's financial home base. Regardless of the currency a rep uses, Salesforce converts all pipeline data into the corporate currency for unified executive forecasting.
    

  • Setup: Open Currency Setup in setup search, click Change Corporate Currency, select your base currency, and click Save.

• Business Hours: 

  • Business hours prevent service automations and SLAs from running during non-working periods. If a customer logs a ticket at 2:00 AM, your response-time clock shouldn't start ticking until business hours open.

  • Setup: Search Business Hours in Setup, set your operating times and time zone, select Active, and click Save.

• Holidays: 

  • Adding company holidays ensures that automated customer support flows, scheduling features, and AI agents recognize when your office is closed.
    

  • Setup: Search Holidays in Setup, click New, enter the date details, and clickSave.

• User Management: 

  • Every employee requires a unique user account paired with specific permissions that govern what data they can view or edit.
    

  • Setup: Search Users in Setup, click New User, fill in personal details, select a license type, and ensure the Username is globally unique across all Salesforce environments.

Setting up multi-currency, corporate defaults, business hours, holidays, and user accounts creates a rock-solid foundation for reporting, security, and automation across your entire company.

When ingesting a Salesforce lead into Data 360, the initial steps are straightforward. You create a data stream, generate a Data Lake Object (DLO), and map it to the standard Lead DataModel Object (DMO). Because a lead belongs to the profile category, you also map it to the Individual DMO to prepare it for identity resolution.

But a profile isn’t just a name; it’s a web of connection points. To truly unlock the power of Data 360, you must also map your lead’s touch points.

What Are Touch Points?

Touch points are the specific channels you use to reach a lead. A single lead profile typically contains multiple touch points, such as:

• Digital: Email addresses and social media handles (X, Instagram, Facebook).

• Traditional: Phone numbers and physical mailing addresses (city, country, postal code).

In Data 360, every touch point has a dedicated, matching Contact PointDMO. For example, emails map to Contact Point Email, phone numbers to Contact Point Phone, physical locations to Contact Point Address, and social handles to Contact Point Social.

Why Must You Map These Touch Points?

Mapping these touch points isn't just about good organization—it is the fuel that powers Data 360’s Identity Resolution engine. Here is why it is critical:

• Accurate Reconciliation: During identity resolution, the system evaluates all available data to merge duplicate records. Well-organized contact points ensure the reconciliation process is flawless.

• Smarter Unified Profiles: The system analyzes your mapped contact points to pick the most reliable, frequent, and trusted touch point, tying it directly to a single Unified Individual ID.

• Simplified Data Architecture: Keeping these data models distributed yet linked allows the system to easily predict relationships and deliver a perfect, 360-degree view of your customer.

Imagine trying to have a meaningful conversation with someone if you could only remember their first name during one chat, their favorite products in another, and their email preferences in a third. You would constantly repeat yourself, miss the mark, and likely frustrate them.

Unfortunately, this is exactly how many companies interact with their customers. Data lives in silos—split between websites, mobile apps, and marketing platforms.

The secret to fixing this? The Unified Individual ID. Here is a breakdown of how it works and why it is the ultimate "single source of truth" for your business.

The Journey to a Single Source of Truth

When a customer interacts with your brand across different platforms, they leave a trail of data. However, to your system, they look like entirely different people. Here is how Salesforce Data Cloud (Data 360) brings them together:

1. The Inflow (Data Streams): Data flows in from multiple sources (like your website, CRM, and mobile app) into a data lake.

2. The Mapping: This data is mapped to a standard data model. At this stage, the customer still exists as multiple separate individual profiles.

3. Identity Resolution: This is where the magic happens. The system uses matching rules to look at these different profiles. If it detects they belong to the same person, it merges them into one Unified Individual Profile.

What Does a Unified Profile Hold?

Once the identity resolution process is complete, everything your customer does is tagged to a single ID. You no longer have to jump between systems to piecemeal their story. Instead, you get a 360-degree view that includes:

• Identity Data: Accurate essentials like their first name, last name, age, gender, and city.

• Behavioral Data: Website actions (tracked via tools like Salesforce Personalization), such as pages browsed, white papers downloaded, or items left in a shopping cart.

• Transactional Data: Purchases and returns made through your mobile app or e-commerce platform.

• Engagement Data: Email interactions (tracked via Salesforce Marketing Cloud), such as which links they clicked, which emails they opened, and what subscriptions they opted out of.

The bottom line: You don't have to look at five different "individuals" to find transaction or behavioral data anymore. You look at one unified profile.

Why This Matters: 3 Big Benefits for Your Business

Using Reconciliation Rules, the system ensures that only the most frequent or highly trusted data makes it into the unified profile. This clean data unlocks three massive advantages:

1. Hyper-Accurate Personalization

When you send a sales email using Marketing Cloud or your CRM, you want to be certain you are using the right details. Unified profiles ensure that basic personalization (like names and locations) and advanced personalization (like product recommendations) are completely accurate.

2. Sharper, Better Segmentation

Data Cloud allows you to build complex filter logic to include or exclude specific audiences. Because all your data sources are mapped to a single unified model, your segments become incredibly precise, helping you target the right people at the right time.

3. Higher Conversion Rates & Zero Duplicates

Duplicate records waste money and ruin the customer experience. By eliminating duplicates and relying on a single source of truth, your marketing and sales teams can interact with customers intelligently, drastically improving engagement and boost conversion rates.

Imagine a customer signs up on your website as "Hers Suri" from "New Delhi." Meanwhile, your CRM already lists them as "Harshit S" from "Delhi." They share the exact same email address, meaning they are the same person. However, because the data is split and slightly conflicting, your system sees two different people.

This is where Identity Resolution comes in. It is a powerful feature within Salesforce Data 360 that merges duplicate records mapped to the Individual Data Model Object into a single, cohesive Unified Individual ID.

To achieve this perfect data harmony, Identity Resolution relies on two critical pillars: Matching Rules and Reconciliation Rules.

Matching rules act as a detective, scanning your data streams to identify which records belong to the same human being. You can leverage two types of matching:

• Exact Match: Looks for identical data, such as the exact same email address or first name ("Hers").

• Fuzzy Match: Catches slight variations or partial data, recognizing that "Suri" and "S", or "New Delhi" and "Delhi", are likely matches.

By combining these (e.g., exact email + fuzzy last name), the system confidently flags the records as duplicates.

2. Reconciliation Rules: Resolving the Conflicts

Once the system groups the duplicate profiles, a new question arises: Which data wins? Should the unified profile say "Delhi" or "New Delhi"? Reconciliation rules resolve these conflicts by defining your preferred values based on three main criteria:

• Most Trusted Source: You can prioritize a specific platform. If your CRM is highly accurate, the system will select the CRM's value ("Delhi").

• Newest Value: The system chooses the most recent data. If the website signup happened today, it will pull the website's value ("New Delhi").

• Most Frequent Value: If the value "Delhi" appears across both a lead and a contact record, but "New Delhi" only appears once on the website, the system picks the most frequent value ("Delhi").

The Audit Trail: Unified Individual Link

Once the heavy lifting is done, how do you track the changes? Salesforce Data 360 provides the Unified Individual Link object. Accessible via the Data Explorer, this object links your new Unified Individual ID back to the original source individual IDs, offering total transparency.

The Bottom Line: Matching rules identify your duplicates, while reconciliation rules solve your data conflicts—giving you a flawless, single source of truth.

In Data Cloud, your data flows from Data Lake Objects (DLOs) into Data Model Objects (DMOs). When dealing with profile data—like Leads from one Salesforce Org and Contacts from another—they all represent the same human being.

To achieve true data unification, Data Cloud uses a core profile DMO called the Individual Object.

• Profile Category Data: Leads, Contacts, and Accounts are mapped here.

• The Goal: Tie different IDs from different systems to a single Individual ID to build a single source of truth.

Syncing Engagement Data from Marketing Cloud

Beyond profile data, you also need to track how customers interact with your brand. This is EngagementData.

By connecting Salesforce Marketing Cloud in Data Cloud Setup, you can deploy Starter Bundles (Email, Mobile Connect, and Mobile Push). This automatically ingests standard data views like clicks, opens, bounces, and subscribers, mapping them instantly into the Data Cloud ecosystem.

Step-by-Step: Mapping Custom Data Extensions

What if you have custom subscriber data inside a Marketing Cloud Data Extension (DE) that isn't part of a standard bundle? You have to map it manually. Here is how:

1. Create a New Data Stream: Select Salesforce Marketing Cloud as the source and locate your specific Business Unit and DE folder.

2. Select the Category & Primary Key: For custom profile data, set the category to Profile and choose your unique identifier (e.g., Subscriber Key) as the Primary Key.

3. Map to Data Model Entities: * Create a Custom DMO to mirror your Data Extension fields.

   • Link and map your Subscriber Key to the Individual DMO (Individual ID) to ensure unification.

   • Map contact details to standard helper DMOs like Contact Point Email, Contact Point Phone, and Contact Point Address.

Once deployed, Data Cloud runs the stream, seamlessly feeding your custom marketing insights into a unified, actionable customer profile!

To start merging your data, you must tell the system which base objects to evaluate.

• Navigate to Identity Resolution and click Create New Rule Set.

• Select the Individual Object as the primary Data Model Object (DMO) you want to unify.

• Define your Rule Set Name (e.g., People) and a unique Rule Set ID.

Once created, the system automatically prepares the target architecture. For every standard object (like contact point email or phone), Data Cloud creates a corresponding unified object, such as Unified Individual and Unified Individual Email.

Step 2: Configure Matching Rules

Next, you define the logic Salesforce uses to catch duplicate records. Because customers rarely provide identical information everywhere, leveraging built-in AI precision rules is essential:

• Fuzzy First Name: Accounts for variations or typos (e.g., matching "Harsh" with "Harshit"). You can adjust the AI precision level to low, medium, or high.

• Exact Last Name: While people play with their first names, surnames usually stay consistent. Setting this to "Exact" keeps your matching accurate.

• Exact Normalized Email: This looks at the core email address, ignoring minor formatting differences.

Data Quality Tip: For normalized matching to work effectively, ensure good data hygiene upfront. Validate customer emails using OTPs or double opt-ins before importing them into Data Cloud.

Step 3: Run and Wait for Reconciliation

After saving your rules, click Update Status. The backend AI will begin processing, running reconciliation rules to link matching IDs across your Salesforce CRM, Marketing Cloud subscriptions, and external data lakes.

Keep in mind that this reconciliation process typically takes 2 to 3 hours to complete. Once finished, you can view the clean, single-source-of-truth profiles in the Data Explorer, fully primed for advanced segmentation and targeted marketing campaigns.