Data expressions in automations

Data expressions let you extract and transform data from one automation step and pass it into the next. Instead of working with raw JSON responses, you can pull out exactly the fields you need — a contact name, an email address, a count of items — and use them downstream in your workflow.

Where you can use data expressions

Data expressions are available in trigger steps and action steps within a regular automation. You can place them in any text field within the step's configuration — for example, an email address field, a message body, or a filter value.

Step type	Supported	Notes
Trigger	Yes	Can reference trigger output data
Action (workflow step)	Yes	Can reference any previous step's output
Goal	No	Goal steps do not process data expressions

info

Data expressions are not available in snippets. Because snippets are reusable fragments that get inserted into different automations, step IDs from the original context would not resolve correctly.

Adding a data expression

When you configure an action step in the automation builder, you can pull in data from a previous step using the Add dynamic content dialog.

1. Open an action step in the automation builder and click into a configurable text field
2. Select Add dynamic content
3. Under Category, choose From a previous step
4. Under From a previous step, select which step to pull data from (e.g., Trigger)
5. Select the output field you want (e.g., Form content)
6. Optionally, enter a path expression in the Extract value field to pull a specific piece of data from the response

The Extract value field uses JQ syntax. If you leave it blank, the entire output of the selected field is used. If you enter an expression like .data.contact.first_name, only that specific value is extracted.

tip

The Extract value field shows example expressions as placeholder text. Common patterns include .data.name, .results[0].id, and .items | length.

Expression syntax

Expressions use JQ syntax — a lightweight language for extracting and transforming JSON data. Here are the patterns you will use most often.

Extract a field

Pull a single value from a step's JSON output.

Expression	Input	Result
.name	{"name": "Alice", "role": "admin"}	Alice
.email	{"email": "alice@example.com"}	alice@example.com

Access nested data

Reach into nested objects or arrays.

Expression	Input	Result
.data.results[0].name	{"data": {"results": [{"name": "Alice"}]}}	Alice
.address.city	{"address": {"city": "Toronto"}}	Toronto

Count items

Use the pipe operator to pass data through functions.

Expression	Input	Result
.items | length	{"items": ["a", "b", "c"]}	3
.contacts | length	{"contacts": [{"id": 1}, {"id": 2}]}	2

Format strings

Combine multiple fields into a single string.

Expression	Input	Result
"\(.first) \(.last)"	{"first": "Alice", "last": "Smith"}	Alice Smith

Work with arrays

Select, sort, and filter array data.

Expression	Description
.[0].name	Get the name from the first item
.items | sort_by(.date)	Sort items by date
.items | map(.name)	Extract all names into a list
.items | map(select(.active == true))	Filter to only active items
.items | unique	Remove duplicate values

Convert types

Change data types when a downstream step expects a specific format.

Expression	Description
.count | tostring	Convert a number to text
.amount | tonumber	Convert text to a number
.data | tojson	Convert an object to a JSON string
.data | keys	Get all field names from an object

Supported functions

Data expressions support a curated set of functions for safe, predictable execution. Here are the most useful categories:

Category	Functions
String	ascii_downcase, ascii_upcase, split, join, ltrimstr, rtrimstr, startswith, endswith, test, gsub, sub
Math	add, floor, ceil, round, min, max, sqrt, pow
Array	length, sort_by, group_by, unique_by, flatten, reverse, first, last, map, select
Type	type, tostring, tonumber, tojson, fromjson, keys, values, has
Object	to_entries, from_entries, with_entries, contains, inside
Format	@base64, @base64d, @html, @uri, @csv, @json

warning

Functions that access the environment or system (env, debug, input, halt) are not available. Expressions are limited to 500 characters. Custom function definitions (def) and imports are not supported.

Graceful error handling

Data expressions are designed to fail safely. If an expression encounters an error — invalid JSON, a missing field, or a syntax issue — it returns the original value unchanged. Your automation continues running without interruption.

This means:

• A missing field returns an empty string, not an error
• Non-JSON input is passed through as-is
• An invalid expression returns the original value

tip

Even though expressions fail gracefully at runtime, we recommend testing your expressions before deploying an automation to production. Invalid expressions are caught during save-time validation so you will see an error if the syntax is wrong.

Use case example

Create a contact from a third-party app via webhook

A third-party app (e.g., a booking platform or lead gen tool) sends a webhook to Vendasta when a new customer signs up. The webhook payload contains nested contact data. Use data expressions to extract the fields you need and create a contact automatically.

Trigger: Webhook received

Example webhook payload:

{
  "event": "new_signup",
  "data": {
    "contact": {
      "first_name": "Alice",
      "last_name": "Smith",
      "email": "alice@example.com",
      "phone": "+14165551234"
    },
    "source": "booking-app",
    "signed_up_at": "2026-04-21T14:30:00Z"
  }
}

Action: Create a contact — use Add dynamic content → From a previous step → Trigger for each field:

Contact field	Extract value
First name	.data.contact.first_name
Last name	.data.contact.last_name
Email	.data.contact.email
Phone	.data.contact.phone

Each expression reaches into the nested webhook payload and pulls out exactly the value needed — no manual data entry, no raw JSON.

Best practices

1. Start simple — Use basic field extraction (.name, .email) before reaching for complex expressions
2. Test in stages — Verify each expression works before combining multiple in one step
3. Use descriptive step IDs — Clear step names make data expression tags easier to read and maintain
4. Stay under the limit — Expressions have a 500-character maximum, so keep them focused on one transformation
5. Combine with logic steps — Pair data expressions with if/else steps for powerful conditional workflows based on extracted data

Learn more

• JQ syntax reference — Full documentation for the JQ query language