Import sets & transform maps in Administration

Import sets & transform maps

Load external data and map it cleanly onto ServiceNow tables.

Quick answer

Key takeaways

The pipeline
Coalesce, insert vs update
Transform scripts and field maps
Scheduling recurring loads
Common mistakes

Import sets are the standard way to load external data into ServiceNow, a spreadsheet of users, a feed of assets, a nightly JDBC pull, and map it onto the right target table without corrupting your clean data. The key idea is a two-stage pipeline: land raw data in staging, then transform it into the target.

The pipeline

Data source            Import Set table       Transform Map         Target table
(file / JDBC / REST) -->  u_imp_xyz (staging) -->  field maps + scripts -->  sys_user
                          raw, messy              coalesce, convert         clean

Raw rows land in a staging table first, so bad or malformed source data never touches your target directly. A transform map then copies each staging column into the correct target field, applying conversions and reference lookups.

Coalesce, insert vs update

A field marked coalesce is the match key. On each row, ServiceNow looks for a target record with the same coalesce value:

coalesce on email:
  match found     --> UPDATE the existing user
  no match        --> INSERT a new user

no coalesce field --> every import INSERTs --> duplicates

You can coalesce on multiple fields (all must match), e.g. first_name + last_name + employee_number.

Transform scripts and field maps

Simple columns map field-to-field. For the rest, use transform scripts:

onBefore

Clean/enrich a row before it writes, split a name, look up a reference. Set ignore = true to skip.

onAfter

Act once the target row is written, create related records.

onStart / onComplete

Run once per import run.

Field map (script)

Compute a single target field from source columns.

Reference fields need care: a source value like a department name must resolve to a sys_id. Use a coalesce/reference lookup in the field map, or the reference will silently import empty.

Scheduling recurring loads

A Scheduled Data Import (a scheduled job) can run a file or JDBC import on a cadence, the standard pattern for nightly HR or asset feeds.

Common mistakes

No coalesce field, so every import inserts duplicates.
Mapping a reference to a display value that doesn't resolve, leaving it empty.
Transforming straight into production without checking the import log/rejects first.
Loading huge files in one transaction instead of batching.

Practice challenge

+0 XPStreak ×0

Question 1 of 3

Imported data first lands in a...

Frequently asked questions

What does the term Import sets & transform maps refer to in ServiceNow?

What is worth remembering about Import sets & transform maps in practice?

Raw rows land in a staging table first, so bad or malformed source data never touches your target directly.

What is another point to note about Import sets & transform maps?

A transform map then copies each staging column into the correct target field, applying conversions and reference lookups.

What tends to go wrong with Import sets & transform maps?

No coalesce field, so every import inserts duplicates. Mapping a reference to a display value that doesn't resolve, leaving it empty. Transforming straight into production without checking the import log/rejects first.