Best Practices for Large Data Volumes and Handling LDV, Skew, and External Objects

In today’s digital world, data is everywhere—created by users every second of every day. As your Salesforce org grows, so does your data. Before you know it, you're managing millions of records, thousands of users, and gigabytes of storage. Without a scalable data model, this growth can lead to performance bottlenecks, slow queries, and sandbox refresh delays.

So how do you stay ahead of the curve? It starts with smart data modeling.

Watch Out for Data Skew

One of the biggest culprits behind poor performance in large data volumes (LDV) is data skew—when too many child records are linked to a single parent. Salesforce recommends keeping child records under 10,000 per parent to avoid issues like:

• Account Data Skew: Happens when too many contacts or opportunities are linked to one account (e.g., an “Unassigned” account). This can cause record locking and sharing recalculations.
• Ownership Skew: Occurs when a single user owns a large number of records. Sharing rule recalculations can slow down operations like deletes or ownership changes.
• Lookup Skew: Arises when many records reference the same lookup value. This can lead to lock exceptions during concurrent updates, this is because you can place lookup fields on any object in Salesforce, lookup skew can create problems for any object within your organization.

Record Locking & Sharing Challenges

When multiple threads update records under the same parent, Salesforce locks both the child and parent records. Even short-lived locks can cause failures if updates overlap. Similarly, changing ownership or sharing settings on skewed records can trigger expensive recalculations across the role hierarchy.

External Objects: A Smarter Way to Handle LDV

Instead of storing all data in Salesforce, consider using external objects via Salesforce Connect. These objects map to data stored outside your org, allowing real-time access without consuming storage. You can use external lookups and indirect lookups to relate external data to Salesforce records—even without Salesforce record IDs.

This approach is ideal when:

• You only need a subset of data at a time.
• You want to avoid syncing and storing massive datasets.
• You need real-time access to external systems.

Best Practices for Scalable Data Models

• Distribute child records across multiple parent records.
• Avoid assigning thousands of records to a single user.
• Use external objects for data you don’t need to store.
• Monitor lookup relationships for potential skew.
• Design with sharing and locking behavior in mind.

Here’s a concise summary of the types of relationships available to external objects in Salesforce:

Relationship Type	Allowed Child Objects	Allowed Parent Objects	Matching Field
Lookup	Standard, Custom, External	Standard, Custom	Salesforce 18-character Record ID
External Lookup	Standard, Custom, External	External	External ID field on the parent external object
Indirect Lookup	External	Standard, Custom	A custom field on the parent with External ID and Unique attributes

---

These relationships allow Salesforce to link external data with internal records, enabling seamless integration and real-time access without duplicating data.