Fields, Relationships and Record Types in Salesforce

Why does data modeling matter?

Data modeling determines how information is stored, linked, and displayed. When set up well, users find what they need faster, reporting becomes easier, and maintenance stays simple. Think of it as the floor plan of your house: when it's right, everything works together.

💡 A comparison

Fields are the rooms, relationships are the doors between rooms, and record types are the interior styles of a room. Together they determine how comfortable you live (work) in Salesforce.

Fields

Fields store data within an object. You have standard fields that Salesforce created for you and custom fields you add yourself for your own processes.

Standard vs. Custom fields

Standard fields

Come with Salesforce objects. Examples: Name, Created Date, Owner.

Custom fields

You add these when you want to capture additional information, like ContractType__c or ProjectPhase__c.

Useful field types (simply explained)

Text, Number, Currency, Percent

Basic data. Use Currency for amounts (with currency settings), Percent for ratios like win probability.

Date and Date/Time

For deadlines, start and end dates, or moments when something happens.

Picklist and Multi-Select Picklist

Limit choices to a fixed list to keep data clean. Only use multi-select when truly needed.

Checkbox

Yes/no question. Handy for simple statuses like "NDA received?".

Email, Phone, URL

Specific validation and convenient click-to-action in the UI.

Formula field

Automatically calculates values based on other fields. For example Expected Revenue = Amount × Win Probability.

Validation rules (brief)

With validation rules you guard data quality: "If X is missing or has an invalid value, show a clear error message." You don't need to write code; you set conditions and a message for the user.

Relationships

Relationships link records together. This builds context: which contacts belong to which account, which opportunities belong to which project, and so on.

Lookup vs. Master-Detail

Lookup relationship

Think of it as neighbors: they know each other, but are independent. Delete the parent record, and the child record usually still exists.

Master-Detail relationship

Think of it as parent-child: the child cannot exist without the parent. Delete the parent, and the children are deleted too. Permissions and ownership flow down from the parent.

Junction objects (many-to-many)

Sometimes you want to link records in a many-to-many relationship. You do this with a junction object: an extra object with two Master-Detail (or Lookup) relationships. Example: a Training can belong to multiple Accounts and an Account can have multiple Trainings. The junction object AccountTraining__c captures the link.

Roll-Up Summary fields

With roll-up summary fields you count, sum, or determine minimum/maximum of child records on the parent. This only works by default with Master-Detail. Examples: Number of Cases on an Account, or Total project hours on a Project.

Record Types

With record types you can support different processes within one object. Each record type can have its own picklist values, page layouts, and business logic.

Different processes, one object

Use record types when you have clear variants within the same object. This prevents unnecessary custom objects.

Page layouts per audience

Show a different layout per record type. This way sales sees only relevant fields, while service sees different information.

Real-world examples

B2C vs. B2B Leads

Use two record types on Lead. For B2B show fields like Company and Number of Employees. For B2C show consumer-specific fields like Date of Birth and Preferred Channel.

Contract variants

One object Contract with record types Subscription, Project, and Support. Each record type has its own picklist values and layout.

Tips for structure and scalability

1. Consistent naming

Use clear, complete names. For API names: Project_StartDate__c instead of PrjStrt__c. This makes reporting and maintenance easier.

2. Only add new fields when needed

Prevent field sprawl. Reuse existing fields and picklists. Document choices in a brief design document.

3. Document and manage

Write a description with purpose and owner for objects and fields. This helps with management, audits, and supports AI features that use context.

Conclusion

With thoughtful fields, proper relationships, and smart use of record types, you build a data model that grows with your organization. The result: better data quality, more efficient processes, and more reliable reporting.

Next steps

Want to discuss your data model or record types? Our team helps you set things up pragmatically and scalably, with an eye for adoption and management.

View all articles →