Skip to main content
Bringing leads in is split into discovery and mutation. Preview first, inspect the result, then import only the records you want.

Import rows

Send rows to Create leads to import many leads into a list in one call.
The response carries the import’s id. A small import finishes inside the request. A larger one runs in the background: follow it with List or get jobs, passing the import’s id, until status is done, failed or cancelled, and read rows_imported, rows_skipped and rows_failed for the counts.
Only rows that would create a new lead count against a workspace’s monthly lead limit. Rows for contacts that already exist, and rows with no email, phone number or LinkedIn profile, do not. Send the same idempotency_key when you retry so a retry never imports twice.

Find new prospects

Prospecting is a preview then import flow.
  1. Run a search. POST /api/v1/public/searches returns a preview and adds nothing to your leads.
  2. Read the results. GET /api/v1/public/searches/{search_id}?lead_list_id=... returns the same preview again.
  3. Import the ones you want. POST /api/v1/public/leads with the search_id and the keys you chose.
Do not skip the preview. An import can create leads, consume credits and change the list. Describe who you want in plain words, or send filters yourself.
type picks the kind of search: The response holds the preview id (use it as search_id), the results, a cursor for the next page and total_matches. Each person has a prospect_key; each place has an id.

Import the results

For a map search send place_ids instead of prospect_keys. lookup_email and lookup_phone find contact details and cost credits, and only when something is found. Searching and importing follow your plan’s monthly prospect allowance; an empty balance answers 402.

Enrichment

Enrichment may skip records that are already complete, lack a usable company signal, exceed a cap, or have no result. A successful HTTP response does not necessarily mean every record was enriched. See Leads for the enrich, verify and research actions. Use small test lists first, store source metadata, and never retry an import blindly after a timeout.