Update matching FileMaker records from CSV without creating duplicates

1. Establish the destination records

The first import exercise adds new contacts. This exercise updates an existing person from a later CSV.

Create a disposable local practice file with a Contacts table and four Text fields: SourceID, FullName, Phone and City. No auto-enter options are needed. Import the baseline as new records and identify the header row as column names. Preserve the IDs as 001, 002 and 003.

Download the baseline CSV ↓

Save a recoverable copy before updating. Do not perform this exercise in a working Contact Journal, Sales Starter or customer file.

2. Decide what a blank means

Download the raw incoming CSV ↓

The incoming file changes Mira’s phone and city, leaves Leon’s phone blank and includes unmatched ID 004. Our policy: update approved existing people, retain a known phone when the source is blank and review new IDs separately.

A direct matching import does not implement that policy automatically. Prepare a reviewed file and keep the raw source unchanged.

Download the reviewed update CSV ↓

Here the retained phone is written explicitly into the prepared file. For recurring imports, use staging and per-field rules, and re-check the destination before applying a batch: a previously retained value could itself become stale.

3. Choose a controlled matching update

1. On the practice Contacts layout, Show All Records. Confirm three records and unique, nonblank SourceID values.
2. Choose File → Import Records → File and select the reviewed CSV.
3. Choose updating matching records. Set SourceID as the match field.
4. Import Phone into Phone and City into City. Leave FullName and unrelated fields untouched.
5. Do not select “Add remaining data as new records.” Identify the header row.
6. Review the mapping and preview. Import, then inspect validation and skipped-record messages.

Matching updates act on the destination found set; a narrow found set can leave an intended contact unchanged. Mapped blanks overwrite values. “Replace existing records” uses sequence rather than the stable-ID match. Review Claris’s import actions before proceeding; update operations are not an Undo workflow.

4. Check every intended value

Show All Records again. Expect three contacts, with no 004 added. All FullName values remain unchanged.

Leon’s number remains because the reviewed file contains it, not because FileMaker ignored the raw blank. Repeat the reviewed update in the same practice state: the count and values should remain the same. This proves the fixture’s repeatability, not universal synchronization.

5. Resolve ambiguous identities first

Duplicate destination IDs make matching ambiguous. Repeated source IDs can apply more than one change to the same person. Blank IDs are not permission to guess from names. Stop and resolve these cases before a real import.

If an expected value did not change, inspect the target occurrence, ID text, found set, mapping and import result. If a phone disappears, inspect mapped blanks. If the record count grows, check whether you chose adding records or adding unmatched rows.

A real file can also have auto-enter rules, validation, privileges and related data absent from this four-field exercise. Inspect those dependencies rather than carrying the practice configuration into production.

6. Keep the batch reviewable

Download the batch checklist ↓

Keep the raw source, approved file, field map, blank-value policy and before/after counts together. Note who resolved uncertain values and when the destination was checked.

Finished result: two intended contacts updated, one existing contact untouched and no extra records. Each changed value has an explanation beyond “the import succeeded.”

References

• Claris: matching import behavior
• Claris: field mapping

Use a practice copy and verify the result in your FileMaker version. Product requirements and platform support remain specific to each product.