Salesforce Data Management and Data Quality
After reports and dashboards show you what is happening in your org, data management is what determines whether you can trust what you are seeing. Effective data management is the cornerstone of a successful Salesforce implementation — but it is also one of the areas most often deferred until problems become visible.
In my experience, data quality issues rarely announce themselves clearly. Duplicates inflate pipeline numbers, incomplete records break automation silently, and inconsistent picklist values make reports misleading without anyone realising why. By the time these problems surface, they have usually been growing for months. Understanding the principles and tools in this chapter early means you can design for data quality from the start, rather than inheriting the cleanup work.
📦 Data Management Best Practices for Salesforce
Section titled “📦 Data Management Best Practices for Salesforce”Managing data effectively is one of the most important skills you’ll develop on the Salesforce platform. Clean, consistent, well‑structured data doesn’t just make reports look better, it directly affects automation reliability, user experience, and the success of every app you build.
🧭 Why Data Management Matters
Section titled “🧭 Why Data Management Matters”Salesforce can be used by many teams concurrently, each with different purpose and expectations. Without clear standards, data quickly becomes inconsistent, duplicated, or incomplete, which leads to broken automation, inaccurate reporting, and frustrated users. As a developer, understanding how data should be created, maintained, and secured is essential to building solutions that scale.
🤔 Core Principles of High‑Quality Data
Section titled “🤔 Core Principles of High‑Quality Data”The most widely referenced core dimensions are: accuracy, completeness, consistency, timeliness, validity, and security.
- Accuracy — The data reflects real‑world truth
- Completeness — All required data is present to do the intended job.
- Consistency — Values do not contradict across systems, records, or time.
- Timeliness — Data is available and up‑to‑date when it is needed.
- Validity - Data conforms to required formats, types, domains, and business rules.
- Security — Access is appropriate and controlled.
Defining what each of these means for your organisation helps ensure reliable reporting and predictable automation.
🏦 Data Governance
Section titled “🏦 Data Governance”Implementing a data governance strategy is fundamental to all data management activities, especially in an enterprise Salesforce environment where the platform may be used by various roles for different purposes. It’s essential to define clear data ownership roles, ensuring that specific individuals or teams are responsible for maintaining, updating, and securing their assigned data.
That being said, you don’t need a full enterprise governance framework to get started. Instead, focus on a few foundational practices:
- Define data ownership — Who is responsible for maintaining specific objects or fields
- Establish standards — Naming conventions, picklist values, required fields, validation rules
- Communicate expectations — Make sure users understand how and why data should be entered
- Review regularly — Use dashboards or reports to monitor data quality over time
These habits create a healthier org and reduce the amount of “cleanup work” developers face later.
Trailhead recommendations: Look at the Data Quality module to discover strategies for assessing and improving the quality of your data in Salesforce.
📚 Core Data Management Practices
Section titled “📚 Core Data Management Practices”Good governance provides the foundation for high‑quality data, but Salesforce also includes powerful built‑in tools that help you enforce those standards day‑to‑day. When governance and platform tools work together, you get data that stays accurate, consistent, and trustworthy as your org grows.
These are the practical, hands‑on habits that keep your Salesforce data clean, reliable, and ready for automation. As a developer, you’ll rely on these practices constantly.
🧪 Implement Validation Rules
Section titled “🧪 Implement Validation Rules”Validation rules verify that entered data meets specified criteria before records are saved. They’re one of the simplest and most effective ways to maintain data quality.
Keep each validation rule focused on one business condition, write an error message that tells the user how to recover, and place it on the relevant field where possible. Test expected saves, blocked saves, integrations, data loads, and any legitimate bypass before deployment. A technically correct rule can still cause a production incident if an integration user or established cleanup process cannot satisfy it.
Common examples include requiring a reason when an Opportunity is closed lost or preventing a status change until the fields needed by the next team are complete.
Trailhead recommendations: Look at the Validation Rules module to learn to maintain data quality in Salesforce by implementing data validation. Then consider the Improve Data Quality for Your Sales and Support Teams project to get hands on experience with validation rules, formula fields, lookup filters, and automation.
🔁 Prevent and Manage Duplicates
Section titled “🔁 Prevent and Manage Duplicates”Duplicate records significantly degrade data quality and create inefficiencies. Implement Salesforce’s built-in duplicate management tools including Matching Rules and Duplicate Rules to automatically identify and prevent duplicate creation. Follow a systematic four-step deduplication process:
- Clean and normalise data before deduplication
- Identify duplicate records using fields like external IDs, email addresses and phone numbers
- Develop clear deduping logic for determining winning records
- Continuously test and iterate your approach
Make deduplication a regular part of your data management strategy with scheduled periodic checks.
Trailhead recommendations: Look at the Duplicate Management module to resolve and prevent duplicate records to increase user confidence in your data.
📥 Data Import and Export Management
Section titled “📥 Data Import and Export Management”Importing and exporting data to Salesforce is often required for maintaining up-to-date and accurate information across your organisation. Whether you’re onboarding new data, migrating from another system, or updating existing records, efficient data management ensures that your Salesforce instance reflects the current state of your business. Importing data allows you to bring in new information, such as leads or customer details, while exporting data enables you to analyse and share insights outside of Salesforce.
Use the Appropriate Tool
Section titled “Use the Appropriate Tool”Choose between Data Loader, Data Import Wizard or third-party tools based on your specific needs.
- Data Import Wizard is an in-built, guided tool for supported standard objects and custom objects. It accepts up to 50,000 records in one import.
- Data Loader supports larger volumes and operations such as insert, update, upsert, delete, and export across a wider range of objects. For very large or recurring movement, assess API and integration patterns rather than treating a desktop import as an operating model.
- Third-party or integration tools can add scheduling, transformation, monitoring, or recovery features. Treat them as production systems: assess security, ownership, failure handling, cost, and support before introducing one.
Import best practices
Section titled “Import best practices”Keep your learning exercises in your practice org. For an authorised production import, confirm approval first, then use a repeatable runbook rather than relying on the operator to remember every precaution:
- Define the change — Confirm the business owner, target object, match key, intended operation, and what must not change.
- Prepare and protect — Export the affected records, retain the approved source file, normalise formats, and validate required and restricted values.
- Control automation and access — Identify Flows, triggers, duplicate rules, validation, sharing, and integration effects before any sample load. Use a dedicated least-privilege operator where possible.
- Test a representative sample — Test normal and edge cases in a sandbox, then reconcile the results before proceeding. For the authorised production import, run and check a small initial batch before loading the remainder.
- Run and reconcile — Save success and error files, compare expected and actual counts, sample the resulting records, and have the business owner confirm the outcome.
- Recover or close — Apply the rehearsed rollback or correction path if validation fails; otherwise store the evidence and remove temporary access.
Trailhead recommendations: Look at the Data Management module to learn how to import and export data in Salesforce. Then follow the Import and Export with Data Management Tools project for hands-on experience with data-loading tools and the Data Import Wizard.
🔐 Security and Compliance of Data
Section titled “🔐 Security and Compliance of Data”Protecting data and meeting compliance obligations are non‑negotiable priorities. Data security is a shared responsibility: Salesforce secures the underlying platform, while your organisation is accountable for safeguarding its own data and complying with relevant regulations. You may work with private and sensitive information, so you must design and enforce strong security controls, and align them with regulatory requirements to reduce legal risk and maintain the confidence of customers and stakeholders.
Implement Least Privilege Access
Section titled “Implement Least Privilege Access”Assign users only the minimum access required for their jobs using a minimal profile, permission sets, permission set groups, and an appropriate sharing model. Regularly review access when people change roles or leave the organisation. Setup Audit Trail can show relevant configuration changes; use access summaries and assignment evidence to determine what a person can actually access.
Code Defensively in Apex and LWC
Section titled “Code Defensively in Apex and LWC”In Apex, treat record sharing and object or field permissions as separate controls. A with sharing declaration enforces record-sharing rules, but it does not by itself enforce object-level or field-level access. Use supported user-mode operations or checks such as Security.stripInaccessible() where appropriate, bind or safely construct dynamic queries, and validate untrusted input on the server. In Lightning Web Components, prefer base components and Lightning Data Service or supported wire adapters, and do not rely on client-side code to protect sensitive business rules.
Monitor Data Access and Changes
Section titled “Monitor Data Access and Changes”Choose monitoring based on the risk and the entitlements available in the org. Field History Tracking, Setup Audit Trail, login history, debug and integration logs, and Event Monitoring each answer different questions; none is a universal audit trail. Name an owner for reviewing the evidence and define what should trigger investigation. Org Health & Monitoring explains when standard field history is sufficient and when Salesforce Shield Field Audit Trail is the better fit for extended coverage and retention.
Trailhead recommendations: Look at the Data Security module to learn about controlling access to data using point-and-click security tools. Then Event Monitoring to discover insights into your Salesforce org with this powerful monitoring feature.
💾 Data Backup and Recovery
Section titled “💾 Data Backup and Recovery”Under Salesforce’s shared-responsibility model, the customer remains responsible for protecting and recovering its data and metadata. Platform availability, the Recycle Bin, and a successful export are not the same as a tested recovery capability.
A recovery strategy may combine native Salesforce capabilities, secure off-platform storage, and a third-party service, depending on recovery objectives and risk.
- Never rely solely on Salesforce’s Recycle Bin, which has limited retention periods and storage constraints.
- Schedule regular backups aligned with your data update frequency and business requirements.
- Test backup and restore processes regularly to ensure data can be recovered quickly when needed.
- Document your backup strategy and communicate it across the organisation for awareness and adherence.
Salesforce’s Data Export Service creates downloadable ZIP archives of CSV data; availability and scheduling vary by edition. Salesforce Backup & Recover and third-party products can add automated data and metadata protection and more granular restore workflows. Choose against documented recovery point and recovery time objectives, retention, storage location, encryption, dependency handling, and the ability to prove a restore works.
Trailhead recommendations: Look at the Backup & Recover Basics to learn how the Backup & Recover solution protects your organisation’s data and metadata.
Salesforce docs on Shared Responsibility: Security Perspective on the Shared Responsibility Model covers data backup explicitly as customer responsibility.
🧹 Data Optimisation and Maintenance
Section titled “🧹 Data Optimisation and Maintenance”Healthy data requires ongoing care and not just one‑time cleanup. Over time, every Salesforce org accumulates outdated records, inconsistent values, and storage pressure. Regular optimisation ensures data stays accurate, performant, and ready to support the automation and analytics a business relies on.
Optimise Data Storage
Section titled “Optimise Data Storage”As your org grows, check data storage and file storage separately. Records such as Cases and Opportunities consume data storage; Salesforce Files and legacy attachments consume file storage.
- In Setup → Storage Usage, review the data and file storage breakdowns to identify what consumes the most space.
- Archive inactive data to reduce storage costs and improve system performance.
- Use those findings to prioritise records, Salesforce Files, and any legacy attachments for archiving under the approved retention policy.
- Establish an approved data retention policy with the relevant business, privacy, security, and records owners.
- Assess an archive design against retrieval, reporting, deletion, security, cost, and restoration needs. Big Objects or an external archive may be part of the design, but neither is a universal answer.
Schedule regular maintenance
Section titled “Schedule regular maintenance”Schedule regular data maintenance activities including monthly audits for duplicates, missing fields, and outdated records. Run regular reports to monitor data quality and identify areas needing attention. Encourage user training on proper data entry practices and maintain clear documentation of all data management processes.
Maintain Documentation
Section titled “Maintain Documentation”Keep detailed records of everything that affects data quality e.g. Validation rules, naming conventions, retention policies, and backup procedures etc. to ensure consistency and support troubleshooting efforts. Good documentation reduces onboarding time, prevents mistakes, and supports long‑term scalability.
By following data management best practices, new Salesforce Admins can establish a solid foundation for maintaining clean, secure, and reliable data that drives business success and user satisfaction.
Trailhead recommendations: Look at the Large Data Volumes module to understand how to work with large data volumes within Salesforce.
✅ Checkpoint: write and rehearse your data plan
Section titled “✅ Checkpoint: write and rehearse your data plan”A backup strategy covers only part of the job. Somebody also has to decide what makes a record trustworthy, what happens to a suspected duplicate, and how work resumes after a damaging change. Bring those decisions into one data-quality, duplicate, backup, and recovery plan that another person could follow, then rehearse its simplest recovery path. Work in your practice org with a handful of invented Accounts and Contacts, name yourself as data owner and recovery operator, and keep any exports somewhere restricted even though the data is synthetic.
📋 Make the four decisions together
Section titled “📋 Make the four decisions together”The example is a contact list used by a support team. Its targets are illustrative business requirements, not Salesforce guarantees; replace them with values you can justify and a tool your org has.
Data quality
Section titled “Data quality”Every in-scope Contact needs an Account and a usable email address. The data owner reviews missing values before each import and weekly afterwards, and assigns exceptions for correction before the list is used.
Keep: the field standards, the check or report used, and before-and-after counts. A format check can’t prove an address reaches the right person.
Duplicates
Section titled “Duplicates”A matching email address makes two Contacts candidates for review, not proof they are the same person. Matching rules find candidates; the duplicate rule decides whether to alert, block, or allow. Name the rules active in your org, and don’t assume a standard rule implements this email criterion exactly. The data owner decides whether to merge and which values must survive.
Keep: one candidate duplicate and one legitimate near-match tested through the real entry route, the observed result, and the decision.
Backup
Section titled “Backup”Example target: lose no more than one working day’s changes (the recovery point objective), with a daily backup and a fresh snapshot before any bulk change. Cover Accounts, Contacts, and the relationship IDs between them, and specify the tool, schedule, retention, storage, access, and who checks for a missed run. Salesforce’s Data Export Service runs at edition-dependent intervals, so it can’t be assumed to meet a daily target; if your tool can’t meet the target, record the gap rather than the wish.
Keep: the latest successful backup time, what it covers and excludes, and a sample you have opened.
Recovery
Section titled “Recovery”Example target: restore the usable contact list within four working hours of the incident being declared (the recovery time objective). The owner approves the scope and the point to recover to; the operator restores, then checks values, relationships, access, automation, and reports before reopening the process.
Keep: the restore steps in dependency order, the reconciled counts, the observed duration, and the owner’s sign-off.
🔄 Rehearse a small, reversible failure
Section titled “🔄 Rehearse a small, reversible failure”A wrong field update is enough to test the plan’s first recovery path without starting from a deletion or a merge. Use a tool that can both export and update, such as Data Loader, and keep its success and error files.
-
Build a known baseline. Create two synthetic Accounts and four Contacts linked to them, each with a Title such as
Service coordinator. Record their IDs and expected values. Keep this set separate from anything you create in the next step. -
Try the two things the plan should catch. Leave one extra Contact’s Email blank and see whether the quality check lets it save; if validation rejects it, record that instead. Then try to create a duplicate of a baseline Contact and note what the duplicate rule does. If it alerts, don’t save anyway. Correct the blank with an invented address such as
alex@example.com, re-run the check, and record both outcomes and the owner’s decision. -
Take the snapshot. Export the two Accounts and four baseline Contacts by ID, including
Id, the Contact’sAccountId, andTitle. Open the files and confirm the values before you change anything. -
Break one thing, then recover it. Note the time, set
Titleon two baseline Contacts toIncorrect test value, then restore just those two values from the snapshot with an update matched byId. Leave every other field out of the recovery file so a repair can’t overwrite unrelated work. Record the elapsed time, and investigate any rejected rows before retrying. -
Reconcile before you call it done. Confirm all four baseline Contacts exist, the two titles match the snapshot, the other two are unchanged, and every
AccountIdstill points to the right Account. Check the report and a representative user’s access. Note any automation or integration effects, or Not applicable with a reason.
This rehearses a field update. Recovering deleted records is a different path: re-inserted records get new IDs, so anything that referenced them needs remapping, as Salesforce’s record-recovery guidance explains; add it to the plan if it is in scope. And a data export is not a metadata backup; configuration, files, and integration side effects need their own cover.
Keep as evidence: the four-part plan, the quality and duplicate results, the snapshot, and the timed recovery-and-reconciliation record, with untested paths marked Not run and given an owner and a date. You have finished when another person can read it and say what is protected, who acts, which recovery path was proven, and what still isn’t. A failed rehearsal counts: record the cause, fix the plan, and repeat that check.
🎯 Final Thoughts
Section titled “🎯 Final Thoughts”Effective data management is less about running the occasional cleanup job and more about building habits that prevent the problems from accumulating in the first place. The clearest sign that a team has this right is that their automation runs predictably, their reports are trusted, and data discussions focus on business questions rather than data quality doubts.
In practice, the orgs where data management is treated as an afterthought tend to share a recognisable pattern: things look fine until they don’t. Duplicates start appearing in dashboards. Automation stops triggering for edge cases nobody expected. A validation rule added too late has to fight years of non-conforming records. These issues are fixable, but they consume time that could have been spent building. Getting the foundations right early, the governance standards, validation rules, duplicate controls, and a clear import strategy, is one of the highest leverage investments you can make in a Salesforce project.
🚀 Next steps
Section titled “🚀 Next steps”Quality data is only valuable if the right people can access it. In Record Access, you’ll explore Salesforce’s record-level security model: Organization-Wide Defaults, the role hierarchy, sharing rules, manual sharing, and the implicit sharing the platform applies on your behalf. You’ll learn how to design secure, scalable access models that support your business without getting in the way of users. The decisions you make here interact directly with the data model and automation you’ve already built.