Flow #9: two touches on LinkedIn before the email

The same email gets a different reply depending on whether the reader has seen your name before. Two small touches are enough to warm up cold email with LinkedIn: a like, then a request, and the prospect moves from “who is this” to “oh, that person”. This flow puts those touches in front of every name on an outbound list, keeps the ones who accept on LinkedIn, and tags the rest for your email tool.

One thing first: Swimlane sends no email. It runs the LinkedIn half and leaves a tag that your automation tool reads.

1. Source List "Outbound this quarter" — a CSV from your CRM, run through "Enrich with LinkedIn"
2. Like their last post Like their last post
3. Connection request Wait 2 days, then a request with no note — wait for acceptance: 5 days
4. Condition Accepted during the request’s 5-day wait?
5. Yes Message Wait 1 day, then one line and one question — this person stays on LinkedIn
6. No Tag Tag "email-ready"
7. No Webhook Outside Swimlane: n8n or Make reads the tag through the API and adds them to your email sequence

Why it works

Our bet, worth testing on your own list: a blank request after a like reads as a natural next step, while a request with a pitch in the note reads as a list being worked. Two light touches beat one heavy one.

The split at the acceptance is the point. The people who accepted are now a LinkedIn conversation; emailing them as well would be noise. The people who did not accept get the email, from a name they have seen twice.

How to warm up cold email with LinkedIn

1. The list. Export the quarter’s accounts from your CRM with a LinkedIn URL column and import the CSV. Then select the list and run “Enrich with LinkedIn”: a CSV contact has no known connection degree, and the request step skips anyone whose degree is unknown. At 80 profiles a day, a 400-name list needs five days before launch.
2. Like, then request. The like runs at once. The connection request carries a 2-day delay, no note, and a 5-day “Wait for acceptance”.
3. The branches. Accepted: a message with a 1-day delay. Not accepted: Add tag “email-ready”.
4. The hand-off, described below, runs in n8n or Make.
5. One channel per person. When someone tagged for email accepts or replies on LinkedIn later, swap their tag for “on-linkedin” from the inbox. The same workflow reads that tag and takes them out of the email sequence.

The warm-up hand-off to your email tool

Swimlane’s webhooks fire only when a contact is created, updated or deleted, and a tag change triggers none. So the hand-off is a scheduled workflow on the public API:

1. Every 15 minutes, call GET /v1/contacts?tag_id=… with the id of “email-ready” (listed by GET /v1/tags).
2. Find the email. A CSV import brings in the LinkedIn URL and the name, not the email, and people found through LinkedIn search rarely come with one. Look each person up in your CRM by LinkedIn URL: that is where the export came from.
3. Add them to the sequence in your email tool.
4. Call POST /v1/contacts/:uuid/tags. It replaces the whole set, so send the current tagIds without “email-ready” and with “in-sequence”. The next run will not pick them up twice.

A second branch does the same for “on-linkedin”: pause them in the sequencer, then clear the tag. The delay is your polling interval, which is fine for email. Setup guides: connecting Swimlane to n8n and connecting Swimlane to Make.

The message

Hi {firstName} — thanks for accepting. I'll keep this short: we help teams like
{company} cut the time between a demo and a signed order form. Is that a problem
you have this quarter, or is it fine as it is?

The question allows “it’s fine”. That is what makes the yes credible.

Guardrails

• Pacing. Requests and messages are paced against your daily ceilings (20 invitations by default), inside office hours. A 400-name list warms up over a month; plan the email launch accordingly.
• The like hits the latest post, however old. Only people who never posted are skipped. If your buyers rarely post, drop the like for that list rather than liking something from two years ago.
• Already connected. Anyone on the list who is already a connection leaves the flow at the request step, with no tag and no message. Filter them out first with the list filter “Relation / Conversation”.
• The “on-linkedin” tag is the fence. Never email someone the flow moved to a LinkedIn conversation.

What to measure

Compare the reply rate of the warmed email sequence with your previous cold one. That difference is the whole return on this flow, and the number worth showing your team.