Skip to main content

Understanding Data Types in Autobound

This article explains each column type (data, action, export), how they behave, and best practices for using them effectively.

Daniel Wiener avatar
Written by Daniel Wiener
Updated over a week ago

Autobound's column fall into 3 groups

Group

Colour

What It Holds

One‑Sentence Rule

Data

Blue

Anything that describes a prospect or company.

Data powers Actions.

Action

Green

AI jobs that create something from Data.

Actions wait for Data.

Export / Send

Purple

Destinations that deliver Action results.

Exports wait for successful Actions.

Group order (Data → Action → Export) is fixed; you can reorder columns inside a group, but never drag a green column left of blue, or purple left of green.


2 · Data Columns in Detail

Data columns are where all prospect-identifying and context-setting information lives. You’ll use these to tell Autobound who someone is and what matters about them.

2.1 Lookup Fields 🔍

Identifiers Autobound needs to find a person or company.

  • contactEmail

  • contactLinkedinUrl

  • contactCompanyUrl

Use when: you need to trigger insights or content generation. These are minimum requirements for any automation.

Tip: Map at least one Lookup field during import. No Lookup = no insights.

2.2 Conversational Accuracy Fields 🧩

Explicit overrides that tell the AI how to talk about the prospect.

  • Contact Name → "Jon" beats "Jonathan".

  • Contact Company Name → "Meta" beats "Meta Platforms, Inc.".

Rules

  1. Can be filled manually or via import/API.

  2. Always override resolved data—even if that data came from enrichment.

  3. Changing a Conversational Accuracy field does not auto‑trigger downstream Actions; regenerate manually when you’re ready.

Use when: you'd like to define the exact name of a user or prosepct in the email, rather than leaving it up to Autobound.

2.3 Added by You Fields

Custom fields appear in the Added by You tab of the Edit Columns modal. They are split into two sub-sections:

Sub‑Tab

What It Shows

Typical Use

Active

Fields that are already mapped and being used. Includes user-created columns and fields referenced from others.

Tags, routing logic, campaign-specific variables.

Available

Fields detected from data imports that are not yet mapped.

CRM fields like HubSpot Lead Source, Lifecycle Stage, SFDC custom properties.

Use when: you need to enrich your campaigns with your own first-party data. These fields do not influence content unless used in a prompt or mapped to an Action manually.


3 · Action Columns in Detail ⚡️

An Action is an AI task linked to a single column—"Find 20 Insights", "Step 1 Email", etc.

States

  1. Waiting… (queued)

  2. Loading… (processing)

  3. ✔︎ Success (result stored)

  4. ✖︎ Error (hover to read why)

Auto‑run toggle in the column header decides whether new rows fire automatically.

Use when: you want to generate something automatically for each row (insights, content, prompts), based on the available Data fields.

Note: Actions may use data from any visible field, but only fields mapped in Insight Control will affect the actual content that gets written.


4 · Export / Send Columns in Detail ✈️

Exports move data out of Autobound.

Examples:

  • Send with Gmail / Outlook – emails from connected inboxes.

  • Send with Outreach / Salesloft – sequences populated via custom fields.

  • Export to Salesforce / HubSpot – write‑back to CRM.

  • POST via API – push JSON anywhere.

Run Modes

  • Auto‑Pilot → fires as soon as every upstream Action is ✔︎.

  • Co‑Pilot → waits for manual review (button reads Review & Send).

Use when: your sequence needs to send content, sync to CRM, or pass data into another system. Configure once per campaign, then let it run.


5 · The Edit Columns Modal—Your Control Room

  • Search bar filters across every group.

  • Icons: 🔍 Lookup 🧩 Accuracy ⚡︎ Action ✈︎ Export.

  • Hide, Delete, Re‑order, or Change Type from the column menu.

  • Custom Fields appear in the "Added by You" section and are grouped as Active or Available depending on whether they are currently used.

Changing a field’s data type is allowed; invalid values stay visible and labelled so you can clean them up.


6 · Walk‑through Example

  1. Import rows with Email + LinkedIn URL → blue Lookup columns appear.

  2. Action "Step 1 Email" drafts copy using those insights.

  3. Export "Send with Outreach" fires once Action is complete

  4. You notice the greeting says Jonathan. Overwrite Contact Name with Jon (🧩). Rerun the email Action → Rerun to resend via Outreach.


7 · Best‑Practice Checklist

  • Use 🧩 Accuracy overrides only when you must—they trump everything.

  • Keep group order intact: Data → Action → Export.

  • Remember: Content is written using values from your Content Hub and fields toggled in Insight Control—not just what’s in your visible table.

Master these rules once, and every future campaign flows from raw data to perfectly personalised outreach—without mystery steps or accidental re‑sends.

Did this answer your question?