What this measurement covers
This design measures how often the option set of a CRM picklist changes without a corresponding migration note. It does not measure whether a change was a good decision, and it does not rank administrators. The useful question is narrower: when a picklist gains, loses, or renames an option, can a later reviewer find a record of who changed it, when, and why? That evidence matters because reports group by option labels, and a silent rename can move a whole segment without changing a single deal.
Define the picklist population
Start with a written list of the picklists in scope. Include the object, the field API name, and the business purpose of each list. Exclude free-text fields and system-generated enumerations that the team cannot edit. Decide whether multi-select fields count, and record the decision. If the CRM exposes a change history for the field, note how far back it goes. The population should be small enough to review in one sitting and stable enough that the same list can be pulled next month.
Choose the observation window
Pick a window that matches the team's release and reporting cycle, such as one quarter. State the start and end dates and the time zone used for the cutoff. If the window includes a known migration, split the analysis so planned changes are not counted as unrecorded ones. Record any period when change history was disabled or incomplete. A shorter window with complete history is more useful than a longer one with gaps, because a missing event looks identical to no event in a count.
Capture the current option set
Export the current options with their labels, internal values, order, and active status. Store the export with a timestamp and the user who ran it. This snapshot is the end point of the comparison. If the CRM allows hidden or archived options, capture those separately, because a hidden option can still exist in historical records and reports. Do not edit the option set during the review. The snapshot should describe the state as it is, not the state the team intends.
Reconstruct the previous option set
Use the change history, audit log, or a previously saved export to reconstruct the option set at the start of the window. If no history exists, state that limitation and use the oldest reliable export you can find. Record how the previous set was reconstructed and how confident you are in it. A reconstruction from memory or from a report's group labels is weak evidence and should be labeled as such. The comparison is only as good as the baseline.
Identify additions, removals, and renames
Compare the two sets item by item and classify each difference. An addition is a new label or internal value; a removal is an option that existed before and is no longer active; a rename is the same internal value with a different label. Keep renames separate from removals, because a rename can silently change how historical records appear in a report. Record the date of each change when the history provides it. If the history shows only the final state, mark the timing as unknown rather than guessing.
Match each change to a migration note
Search the team's change log, release notes, and ticket system for a record that mentions the field or the option. Match on the field API name and the option label, not only on a project name. Record whether a note exists, who wrote it, and whether it states the reason and the reporting impact. A note that says "updated picklist" without naming the option is weak evidence. The measurement should distinguish a specific note from a generic mention.
Quantify unrecorded changes carefully
Report the number of changes, the number with a matching note, and the number without one. Report additions, removals, and renames separately, because they carry different risk. Do not present a single "compliance rate" as if every change were equally important. A renamed option used by an executive dashboard deserves more attention than an added option that no report references. Keep the record-level table so a reviewer can see which changes drive the result.
Assess reporting impact
For each unrecorded change, check whether any report, dashboard, filter, or automation references the option. Record the report name, owner, and whether the change would alter a total or a segment. This step turns a documentation gap into a risk statement. A picklist change with no downstream use is a bookkeeping issue; a rename that shifts a segment in a board report is a reporting defect. The distinction should shape the repair priority.
Look for patterns across fields
Group the results by field, by administrator, and by release period. Look for a picklist that changes repeatedly without notes, or a period when several fields changed together, which may indicate an undocumented migration. Patterns are more actionable than individual misses, because they point to a process step rather than a person. Do not use the pattern to assign blame; use it to decide where a required note or approval would prevent the next gap.
Choose one control and retest
Pick one change type and add a control. That might mean requiring a ticket reference before a picklist edit, adding the field to a release checklist, or generating a weekly diff of option sets. Retest with the same population and window definitions. Compare the new result with the baseline and report what changed. A control that reduces unrecorded renames without slowing routine additions is a real improvement, even if the total number of changes stays the same.
Document limits and data handling
This measurement depends on the completeness of the CRM's change history, which may be short or unavailable. It cannot prove intent, and it does not capture changes made directly in the database. State those limits in the method note. The extracts may contain customer data if option values include names or segments, so keep only the fields needed and restrict access. Delete the working extracts when the review is complete. A repeatable review is one that another administrator can run without reconstructing the method from memory.
Sources
- https://www.nist.gov/cyberframework
- https://www.census.gov/data/developers/guidance/api-user-guide.html
- https://knowledge.hubspot.com/records/understand-original-source-properties
- https://www.w3.org/TR/NOTE-datetime
- https://www.ftc.gov/business-guidance/privacy-security