Integrations
Instantly n8n Integration
We build n8n pipelines against the Instantly API that turn a plain-English ideal-customer description into a scored, verified contact list, and stop before anything is sent so a person can review it.
Turning a sentence like "mid-size e-commerce ops teams in the Nordics" into an actual list is three separate jobs: find the companies, pick the right person inside each one, and get a deliverable address for that person. The pipeline does all three and then hands the result back for review.
What we build
Six stages, in order. The description goes in at the top and a working list comes out at the bottom.
- 1
ICP to queries
A language model expands the plain-English ideal-customer description into concrete search queries and Instantly filter parameters, so the operator does not have to guess at boolean syntax.
- 2
Company discovery
Candidate companies are gathered from those queries and deduplicated before anything downstream touches them.
- 3
Fit scoring
Each company is scored against the ICP before any contact work starts, and the score carries the reasoning that produced it.
- 4
Decision-maker identification
The pipeline looks for the role that owns the problem being solved, rather than whoever happens to be listed first at the company.
- 5
Contact discovery and verification
Email addresses are found and validated, because a list that bounces damages the sending domain it was built for.
- 6
Google Sheets delivery
Rows land in the sheet the operator already works in, with the evidence behind each row visible so it can be checked or deleted.
Two design decisions
Score before enriching
Enrichment and verification are billed per record. Scoring the company against the ICP first means that budget is spent on companies that already passed the check, not on a raw scrape.
The pipeline does not send
It produces a candidate list with its reasoning attached and stops there. A person reviews before anything goes out, which is the whole point of putting the review step inside the automation instead of hoping someone remembers to do it.
What it connects to
n8n is the orchestrator. Three systems hang off it, and each one has a single job.
Instantly API
Called directly from n8n over HTTP with an Instantly API key credential. This is where the filter parameters, the lead records, and the campaign side of the work live.
Google Sheets
The default delivery target. One row per contact, with the fit score and the evidence for it in adjacent columns so the review step is a read, not an investigation.
A language model
Used for two jobs only: expanding the ICP description into queries and filter parameters, and scoring each discovered company against that description.
FAQ
Do you maintain an Instantly n8n node?
No. The n8n-nodes-instantly package is maintained by Instantly. We build workflows against the Instantly API directly, and we publish a different node, n8n-nodes-apollo, for Apollo.io.
Apollo integration pageDoes the pipeline send the emails?
No, and that is deliberate. It ends with a reviewable list. Sending is a separate decision made by a person looking at the rows, and connecting the reviewed list to a campaign is a second, explicit step.
What do you need from us to build it?
An Instantly account with API access, a language model API key, and a Google account for the destination sheet. Plus the part that actually matters: a written description of the customer you want, specific enough that a person could sort a list with it.
Can the list land somewhere other than Google Sheets?
Yes. The sheet is the default because it is where review already happens, but the delivery stage is one node. Anything n8n can write to works as a target, including a CRM or a database, once you have decided who reviews the rows and where.
Want a list you can actually review?
Describe the customer you want in a sentence. We will scope the pipeline that turns it into rows.