A trigger or action produces output fields; the next step in your flow needs input fields — and those two field lists are almost never identical in name, shape, or format. Data Mapping is where you tell the platform exactly how one step’s output becomes the next step’s input, and it’s a step you’ll touch in essentially every flow you build.
Every step after the first in a flow has a mapping screen showing the incoming fields on one side, sourced from a real data sample pulled from the previous step, and the fields the current step expects on the other. Mapping a simple value is a matter of connecting the two — drag a source field onto a destination field, or type a fixed value directly into it.
Source and destination fields don’t always line up cleanly — a date might need reformatting, a full name might need splitting into first and last, a numeric total might need to be summed across an array. For that, the mapper has a developer mode where any field’s value can be written as a JSONata expression instead of a straight connection — the same expression language used in Content-Based Routing Overview conditions. The mapper evaluates the expression against your real data sample as you write it, so you can see the actual result before the flow ever runs.
Mapping gets more interesting once either side involves an array — for example, mapping each line item of an order to a corresponding line in an invoice. The platform supports array-to-array mapping directly in the same interface, including cases where the array itself contains objects with their own nested fields, without requiring a separate step or a hand-written loop.