Salesforce Data 360 maintenance checklist and guide

Salesforce Data 360 (formerly Data Cloud) is now the enterprise data backbone for AI, CRM, and cross-cloud activation. Letting it run on autopilot is a fast path to runaway costs, degraded profiles, and stalled AI workflows.

Whether you're an IT admin, a data analyst, or a business stakeholder, this guide walks you through the essential Data 360 maintenance tasks you should be doing daily, weekly, monthly, and quarterly, plus a ready-to-use checklist you can hand off to your team.

Why Data 360 maintenance matters more than you think

Data 360 isn't a set-and-forget platform. Unlike traditional Salesforce admin tasks, maintenance here has a direct financial impact. The platform runs on a consumption-based credit model, meaning every data ingestion, identity resolution run, calculated insight refresh, and segment publication draws from your credit pool.

Key risks of neglecting maintenance:

• Unexpected credit overruns from unmonitored scheduled processes
• Profile degradation when identity resolution rules go stale or conflict
• Silent ingestion failures that don't surface until downstream activation breaks
• Governance drift as new data sources get added without proper classification or consent mapping
• Schema mismatches that cause ingestion pipelines to fail or corrupt data silently

Who should own Data 360 maintenance?

Maintenance responsibilities typically span three roles:

RoleMaintenance focusSalesforce/Data AdminMonitoring, access control, credit usage, setup audit trailData Architect / AnalystIdentity resolution, data modeling, ingestion schema, calculated insightsBusiness / Marketing OpsSegment health, activation performance, consent and compliance

For smaller teams, one person may cover all three. The checklist below is organized so each role can quickly identify what belongs to them.

Data 360 maintenance checklist

✅ Daily checks

For admins:

• Check ingestion pipeline status; confirm all data streams ran successfully and flag any failures
• Review login history for unexpected access or failed attempts
• Monitor Salesforce Trust (trust.salesforce.com) for any platform incidents affecting your org or Marketing Cloud connection

For data teams:

• Spot-check unified profile counts in Profile Explorer for unexpected spikes or drops
• Confirm real-time ingestion APIs are streaming correctly (if using streaming ingestion)

✅ Weekly checks

Credit and usage monitoring:

• Open the Digital Wallet / Data Services Consumption Card and review credit burn rate against your expected pace
• Identify which processes consumed the most credits, with focus on Identity Resolution (highest multiplier), Calculated Insights, and Segment Publishing
• Check for any automated processes that ran unexpectedly (e.g., editing an Identity Resolution ruleset triggers an automatic full refresh; confirm this was intentional)

Identity resolution:

• Review match rate trends; a sudden drop may signal a schema change or a data source issue upstream
• Confirm no unintended ruleset edits were saved (these trigger automatic re-runs and credit consumption)

Data ingestion:

• Audit data stream schedules; remove or pause any streams that are no longer feeding active use cases
• Validate that delta/incremental loads are working correctly; unexpected full-load runs inflate credit usage significantly

Access and governance:

• Review Setup Audit Trail for any configuration changes made during the week
• Confirm no unauthorized users have been granted Data 360 permission sets

✅ Monthly checks

Data model and schema hygiene:

• Review Data Lake Objects (DLOs) and Data Model Objects (DMOs) for orphaned or unmapped fields
• Confirm all schema definitions are versioned in your source control system; schema mismatches cause ingestion failures or silent data corruption
• Audit naming conventions; Data 360 enforces strict object and field naming rules and non-compliant names can break pipelines

Identity resolution review:

• Re-evaluate match rules against current data volume and use cases; rules that made sense at launch may be over- or under-matching as data grows
• Review reconciliation rules to ensure field priority (e.g., which source "wins" for email address) still reflects your business intent
• Validate unified profile counts are stable and not drifting due to stale or conflicting source data

Segment and calculated insight audit:

• Review all active segments; archive or delete segments that are no longer used in any activation
• Check Calculated Insight refresh schedules; remove insights that no one is consuming downstream
• Confirm segment publish schedules align with actual business cadence (over-publishing is a common credit drain)

Consent and privacy:

• Verify consent data is current and properly mapped across all active data spaces
• Review AI-driven PII tagging (introduced in Summer '25) to ensure sensitive fields are classified correctly under your GDPR/HIPAA policies
• Audit data space configurations; confirm each data space is scoped to the right brand, region, or department

✅ Quarterly checks

Strategic and architecture review:

• Assess your org provisioning strategy; if you've added new Salesforce orgs, evaluate whether Data 360 One (the recommended multi-org approach) or Zero Copy federation is the right model for sharing data
• Review your mix of batch vs. real-time ingestion; real-time is powerful but more expensive, so confirm the use cases justify the cost
• Evaluate whether unstructured data sources (text, images, external datasets) are being handled by semantic modeling, not manual workarounds

Release and certification maintenance:

• Review Salesforce release notes for any Data 360 changes (Salesforce releases updates three times per year: Spring, Summer, Winter)
• If your team holds Data 360 Consultant certifications, complete any required maintenance modules on Trailhead before the deadline; certifications now require active renewal with each major release
• Test any new features in a Data 360 sandbox before deploying to production; use sandbox migration to validate configurations first

Cost optimization review:

• Compare credit consumption against the three main buckets: Connect and Harmonize, Analyze and Predict, and Act; identify which bucket is disproportionately high
• Consider whether a "stop consumption" flow is needed; you can use Salesforce Flow and the Connect API to automatically pause data stream schedules, calculated insight refreshes, or segment publishing when credit thresholds are hit
• Revisit your credit purchasing model (pre-purchase vs. flex vs. pooled) with your Salesforce AE if usage patterns have shifted significantly

Key takeaways

Data 360 is a living platform. Every ruleset edit, new data source, and segment addition changes what your platform costs and how it performs. Structured maintenance isn't optional; it's how you protect your investment.

Here are the five most impactful maintenance habits to build:

1. Watch Identity Resolution like a hawk. It carries the highest credit multiplier; saving, editing, or triggering a refresh can consume tens of thousands of credits in a single run. Any change should go through a change-management process before touching production.
2. Audit your segment and insight schedules monthly. Stale segments and over-scheduled calculated insights are the most common source of unnecessary credit burn.
3. Version-control your schema definitions. Schema files act as contracts between source systems and Data 360. A mismatch can cause silent data corruption and you won't always know until a downstream campaign or AI workflow fails.
4. Use sandboxes before every production change. The Data 360 sandbox (introduced in Winter '26) lets you test identity resolution configs, data stream changes, and package deployments before they hit live data.
5. Set up a consumption alert and stop-flow. The Digital Wallet sends alerts but doesn't auto-stop processes. Build a Flow that automatically pauses scheduled processes when you hit a credit threshold; this is your circuit breaker.

Quick reference: what triggers unexpected credit consumption

ActionWhy it costs more than expectedEditing an Identity Resolution rulesetAutomatically triggers a full refreshRunning a Calculated Insight for the first timeProcesses all rows, not just new onesIncreasing data volume in a connected DLONext refresh processes more rows than estimatedAdding a new match rule mid-cycleTriggers a re-evaluation across all existing profilesPublishing a segment more frequently than neededEach publish run draws activation credits

Salesforce Data 360 has matured significantly, evolving from a marketing-focused CDP into the enterprise data and AI foundation across the entire Salesforce stack. That evolution brings enormous capability, but also more operational complexity than a typical Salesforce org.

The teams that get the most value from Data 360 aren't just the ones with the best data model. They're the ones with the most disciplined maintenance habits: monitoring consumption consistently, treating identity resolution changes like code releases, and keeping their schemas and governance policies aligned with the business.

Use this checklist as your starting point. Customize the cadence to your org's size and data velocity and revisit it every time Salesforce drops a major release.

‍