WHY SALESFORCE FIELD USAGE SHOULD BE REVIEWED BY RECORD TYPE.

A field has a 2% fill rate across the whole object. By the global number, it looks unused — an obvious candidate to hide or remove. Then you split the same field by record type and find it is 85% populated on one of them.

Nothing about the field changed. The way you measured it did. That single field is essential to one business process and irrelevant to the rest, and the global fill rate flattened both realities into a misleading 2%.

This is one of the most common ways field cleanup goes wrong: global fill rate alone creates false cleanup candidates. This guide explains how record types change the meaning of field usage, why that matters during cleanup, when record-type review applies, and how to use record-type context to reduce false positives before you make a decision.

Global fill rate compares how many records have a value in a field against the total number of records on the object. It is a strong first-pass signal because it quickly separates fields that are widely populated from fields that are rarely touched.

That is exactly what you want at the start of a cleanup review: a way to narrow hundreds of fields down to a shortlist worth investigating. The limitation is that a single global percentage averages across every record on the object, regardless of which process created it. When an object serves several distinct processes, one number cannot represent all of them faithfully.

Global fill rate is a good starting point. It becomes misleading the moment one object is doing several different jobs.

Record types are how a single Salesforce object supports more than one business process. The same object can carry several record types, each with its own page layouts, picklist values, and the fields that process actually needs.

In practice, record types often map to meaningfully different things:

• Different sales motions or deal types on Opportunity.
• Different account types — customer, partner, vendor, member.
• Different case types — support, billing, escalation.
• Different regions, product lines, or business units.
• Different member or customer categories on custom objects.

Because each record type represents a different process, a field's usage only makes sense in the context of the record type it belongs to. A field that is core to one process will naturally look rare when diluted across every other process on the object.

Consider a single custom field on the Account object, viewed first globally and then by record type:

Globally, this field looks like a clear cleanup candidate at 2%. By record type, the picture is completely different. The field is part of an active process on Member Company accounts and genuinely unused on Prospect and Vendor accounts.

The right decision is not "delete" or "keep" — it is "this is a Member Company field." That framing changes who you talk to, which layouts you check, and what a safe change would even look like. The global number alone would have pointed you toward removing a field that one team relies on.

Acting on global fill rate without record-type context introduces specific, avoidable risks:

• Removing or hiding a field that is used by one process while it looks unused overall.
• Confusing one team's valid field with another team's unused field on the same object.
• Breaking reports, list views, or user workflows that depend on a process-specific field.
• Misclassifying a process-specific field as a global cleanup candidate.

Each of these is hard to catch after the fact, because the team that relied on the field may not notice until the process it supports runs again. Reviewing by record type is a low-cost way to catch the problem before the change, not after.

Inherited orgs are where record-type context matters most, because the history behind the configuration is usually missing.

Over years of changes, inherited orgs accumulate old record types that are no longer obvious, processes that were renamed without renaming the underlying record type, legacy fields that were hidden rather than removed, and inconsistent decisions made by a series of admins who each had their own conventions. A field tied to a record type from a process that was rebranded two years ago can look like noise — until you find the team still using it.

Reviewing usage by record type helps rebuild that missing context. For the wider first-pass approach to taking over an unfamiliar org, see the guide to inheriting a Salesforce org.

Record-type fill rate is one signal among several. Before acting on a low-fill field, work through the full set of usage signals:

For the full reasoning behind why no single usage number is enough, see Salesforce field cleanup: why fill rate alone is not enough, and the broader field cleanup checklist.

Record-type segmentation is valuable, but it does not apply everywhere. It is worth knowing when the signal is not available or could mislead:

• Objects without record types — there is nothing to segment.
• Objects with only one active record type — the breakdown matches the global number.
• Fields that are not queryable or aggregatable for usage measurement.
• Formula and encrypted fields, where populated-value measurement does not behave the same way.
• Low record-count segments, where a small denominator makes percentages volatile and easy to misread.

In these cases, lean on the other signals — layout exposure, references, integrations, and business-owner confirmation — rather than reading too much into a record-type percentage that cannot carry the weight.

Field & Object Audit adds record-type usage context for cleanup review candidates where record types exist. This helps admins and consultants distinguish globally low-fill fields from fields that may matter to a specific process or team.

The record-type view is presented as a field usage signal for review, not as proof that a field is safe to delete. It is one part of a review-ready workbook designed to reduce false positives, alongside the cleanup candidates, supported reference signals, and validation notes you would work through before any change.

Global fill rate is where review starts, not where it ends. On objects that serve multiple processes, a single percentage can quietly hide a field that one team depends on.

A Salesforce field is not unused just because its global fill rate is low. Review by record type before making cleanup decisions.