Why the source label matters
A consent record without a source is hard to defend and hard to use. The source label tells the team where the contact came from, which form or event collected the consent, and which language version was shown. During an import, that label can be overwritten by a default, truncated by a field limit, or replaced by a generic value such as "import." When that happens, the consent may still exist in the source system, but the CRM can no longer explain it. This test checks whether the label survives the import intact.
Define the import and the label under test
Start with one import and one source label. Record the source system, the label value, the field that carries it, and the expected mapping in the CRM. Decide whether the test covers a single label or a controlled set, and write the rule. If the import changes several fields at once, isolate the source label so the result is attributable. The test should be specific enough that a second reviewer can repeat it with the same file.
Capture the raw import file
Save the source file exactly as received, with a checksum and the import date. Record the column that holds the label and a few sample values. Do not edit the file before the test. If the import tool applies a template or a default, record that configuration. The raw file is the evidence of what was sent; the CRM is the evidence of what was stored. The comparison between the two is the test.
Record the pre-import CRM state
If the contacts already exist in the CRM, capture their current source label and consent fields before the import. If they are new, record that the field is empty. Save the record identifiers so the same records can be checked afterward. A pre-import snapshot distinguishes an overwritten value from a value that was never set. It also shows whether the import created a duplicate instead of updating the existing record.
Run the import and inspect the result
After the import, open the same records and record the stored source label, the consent value, and the timestamp. Compare them with the raw file. A label that is present and identical is a pass. A label that is empty, generic, truncated, or replaced is a failure. Record the exact stored value, not only a pass or fail. The stored value is what a future reviewer will see when they try to explain the consent.
Test the update path as well as the create path
Imports often behave differently for new records and existing ones. Test both. For an existing record, confirm whether the import overwrote a more specific label with a generic one. For a new record, confirm whether the label was stored or dropped. Record the result for each path. An import that preserves labels on new records but overwrites them on existing ones is a partial pass, and the distinction matters for the repair.
Check field limits and character handling
A label can be truncated by a field length limit or altered by character encoding. Test a label near the limit and one with punctuation or non-English characters. Record the stored value and compare it with the source. A truncated label may still look plausible while losing the part that identified the source. Include this check in the test rather than discovering it through a failed audit later.
Test duplicate and merge behavior
If the import creates a duplicate, the source label may attach to the wrong record or split across two. If the CRM merges duplicates, the surviving record may keep one label and discard the other. Create a controlled duplicate and run the import or merge, then record which label survives. This is a common hidden loss, because the consent record still exists but the explanation is now attached to the wrong person or lost entirely.
Classify the outcomes
Group the results into label preserved, label overwritten, label truncated, label empty, label generic, and duplicate split. Add a category for a label that survived but is attached to the wrong record. The classification separates a mapping defect from a merge defect and gives each finding a clear owner. A test that only checks whether the field is non-empty will miss the overwrite and duplicate cases.
Measure the survival rate
Report the number of records tested, the number with a preserved label, and the number in each failure category. Report counts alongside percentages and show examples of the stored values. Do not describe the result as an industry benchmark or a legal standard. It describes this import at this time. A small sample with a clear overwrite pattern is more useful than a large sample averaged across different import types.
Choose one repair and retest
Pick the largest failure pattern and fix it at the source. That might mean locking the field during import, mapping the correct column, widening the field, or changing the duplicate rule. Retest the same records and the same label values. Compare the result with the baseline and report what changed. A repair that preserves labels without blocking legitimate updates is the goal.
Document limits and data handling
The test depends on access to the source file and to the CRM records after import. It cannot prove that the label is accurate in the source system, and it does not replace a legal review of the consent itself. State those limits in the method note. Use test records or a controlled sample, restrict access to the files, and delete the working extracts when the review is complete. A repeatable label test is one that another administrator can run before the next import.
Sources
- https://www.ftc.gov/business-guidance/privacy-security
- https://www.ecfr.gov/current/title-16/chapter-I/subchapter-C/part-316
- https://www.w3.org/TR/NOTE-datetime
- https://www.nist.gov/cyberframework
- https://www.census.gov/data/developers/guidance/api-user-guide.html