Three ways to start an import
You can pull Outreach prospects into Twain from three places. The tag picker and field-mapping form are the same in all three — what differs is when the import runs and how often.
Campaign creation. Pick Outreach as the lead source while setting up a new campaign. Prospects on the selected tag land in the campaign immediately and Twain starts research right away.
Leads import (existing campaign). Open a campaign you've already built and click Import leads. Good for topping up an audience, expanding to a new tag, or backfilling a campaign that ran out of leads.
Workflow. Drop an Outreach node into a workflow and the import keeps running automatically. Every minute Twain checks the tag for new prospects, pulls them in, runs research, and drops them into the campaign. No manual step required.
Most teams start with a one-off import and move to a workflow once the campaign is calibrated.
Campaign creation. Pick Outreach from the integrations dropdown to start an import.
Tags are the selector
Outreach doesn't have static lists like HubSpot or Amplemarket. The unit of selection is a tag. Whichever tag (or tags) you pick in the import step is the audience — Twain pulls every prospect carrying that tag in Outreach.
A few things to know:
The tag must exist in Outreach first. Twain reads tags, it doesn't create them. Set the tag on the prospects you want imported, then pick it in Twain.
Re-imports are safe. Twain deduplicates on Outreach prospect ID, so prospects already in the campaign won't be added twice. New prospects you tagged since the last run will be picked up.
Map fields to variables
After picking the tag, you map Outreach prospect fields to the variables Twain uses for research and sequence generation. The defaults cover the basics: work email, first name, last name, job title, LinkedIn URL, company.
LinkedIn URL is worth calling out: it's the field that unlocks deep research. Without it Twain can still draft a sequence, but with it the research is substantially richer because the assistant has a profile to read.
Beyond the defaults, anything you keep on the Outreach prospect is mappable — persona fields, last contacted date, stage, territory, custom fields. The more context you pass at import time, the less generic the output. Variables you map once stay attached to the prospect, so day-one sequences and day-fourteen follow-ups both have the same context.
Mapping step. Pair each Twain variable with the Outreach prospect field that should fill it.
One-off import vs. continuous feed
A campaign-creation or leads-import run is a snapshot: Twain grabs prospects on the tag at that moment, runs research, and stops. Add a new prospect to the tag tomorrow and Twain won't notice unless you re-import.
A workflow import is continuous: Twain re-checks the tag on a polling schedule, pulls anything new, deduplicates, and runs the full pipeline automatically. This is the right setup for ongoing motions — new-hire tagging, intent triggers, territory rebuilds — where the audience keeps growing and you don't want a manual step in between.