A LinkedIn connection then email sequence is a two-channel opener: a connection request with a short note on day 0, then a cold email on day 3 that references the request whether or not it was accepted. The email benefits from name recognition and the request benefits from the email's detail. Teams report 30 to 45% connection acceptance with a specific note and 6 to 11% email replies on the follow-through, provided the email address was found and verified separately; LinkedIn does not supply it.
| Day | Channel | Message | Rule |
|---|---|---|---|
| 0 | Connection request with a 300-character note | One specific reference, no pitch | |
| 3 | Opener that references the request | Works whether accepted or not; never says "since we are connected" unless true | |
| 5 | LinkedIn (if accepted) | One message, two lines, pointing to the email | No pitch in the DM; the email carries it |
| 8 | First follow-up, same thread | The day-3 follow-up template shape | |
| 15 | Second follow-up with evidence | See the second follow-up | |
| 24 | Break-up | The break-up template |
LinkedIn messages stay sparse. Automation on LinkedIn at volume gets accounts restricted; the volume lives in email, where the list can be verified and the domain warmed.
A cold email from a name the reader has never seen gets triaged. The same email from a name they saw in a connection request three days ago gets opened, because the brain files it as a continuation. The lift is in open and reply behaviour, not in anything the LinkedIn note said.
The connection note cannot carry the pitch; 300 characters is not enough and a pitch in a request gets declined. It carries one specific reference: a post they wrote, a talk, a shared group, a mutual connection. The email, three days later, does the work.
The address is the catch. LinkedIn does not give you the email, and the finders that enrich LinkedIn profiles are 75 to 85% accurate at best. Find it, verify it, and send only to confirmed-valid; a bounced email after a connection request is a visible failure to a person who now knows your name.
{first_name}, your post on {post_topic} matched something we see at {sender_company} every week. Not pitching; I would like to follow what you write on it. {sender_first_name}
{post_topic} The specific thing they wrote or spoke aboutLength Count characters; LinkedIn truncates at 300Subject: {post_topic}, continued
Hi {first_name},
I sent a connection request on LinkedIn a few days ago about your post on {post_topic}. Writing here because email is better for the actual point.
The point: {problem} is the thing behind what you described, and {sender_company} handles it for {customer_type}: {outcome_one_line}. {similar_company} wrote about the same issue before they used us; afterwards, {result}.
If it is worth 15 minutes, pick a slot. If not, a one-word reply closes it and I will still read your posts.
{sender_name}
{sender_title}, {sender_company}
If this is not relevant, reply "no" and I will not write again.
{problem} The problem behind their post{result} One numberThanks for connecting, {first_name}. I sent a short email on {post_topic} to your work address on {day}; it has the detail that would not fit here. No need to reply twice.
{day} The day the email wentExport the profile, find the address through a pattern finder or the LinkedIn email finder workflow, then verify it. Catch-all and unknown results go LinkedIn-only; the email leg is skipped.
Twenty to forty requests a day with notes written per person. Bulk automation on profile pages restricts the account and ends the channel.
"Since we are connected" to someone who ignored the request is caught instantly. The email references the request, not an acceptance.
Warmed secondary domain, plain text, one link, bounce under 2%, complaints under 0.3%. The LinkedIn leg is small and manual by design.
| Metric | Healthy range (industry) | If missed |
|---|---|---|
| Connection acceptance | 30 to 45% | The note pitched, or referenced nothing specific |
| Email reply rate | 6 to 11% per 100 delivered | The email repeated the note instead of extending it |
| Lift versus email-only | 1.5 to 2x reply rate | The email went before the request, or more than a week after |
| Bounce rate on the email leg | Under 1% verified | Finder output sent unverified; 10 to 25% of it is invalid |
Subject: Catch-all domains, continued Hi Sofia, I sent a connection request on LinkedIn a few days ago about your post on catch-all domains. Writing here because email is better for the actual point. The point: catch-all acceptance is the thing behind the "verified but bounced" rows you described, and we label it separately so those rows never reach a cold sequence. Bramble wrote about the same issue before they used us; afterwards, their accept-all bounces went to zero. If it is worth 15 minutes, pick a slot. If not, a one-word reply closes it and I will still read your posts. Jess Carr BounceZero If this is not relevant, reply "no" and I will not write again.
Three hundred characters is not enough, and a pitch gets declined. One specific reference.
"Since we are connected" is checked in one click. Reference the request, not an acceptance.
Accounts get restricted. Twenty to forty hand-written requests a day.
LinkedIn gives no address. The finder's guess is 15 to 25% wrong; verify or skip the leg.
The full set
Profile to verified address
The LinkedIn leg in depth
Search, export, enrich, sequence
Four openers for the email leg
How specific the reference needs to be
The request first, the email three days later. The request plants the name; the email arrives as a continuation and gets opened. Reversed, the email is cold and the request looks like a chaser.
One specific reference to something they wrote, said or did, and no pitch, in under 300 characters. Requests with a pitch get declined; requests with a specific reference are accepted 30 to 45% of the time.
LinkedIn does not provide it. Use a pattern finder or an enrichment tool on the name and company, then verify the result; finders are 75 to 85% accurate on LinkedIn-sourced contacts and the rest bounce. Skip the email leg for catch-all and unknown results.
Not at volume. Automated requests and messages on profile pages get accounts restricted. Keep LinkedIn to 20 to 40 hand-written requests a day and put the volume in email, where the list can be verified and the domain warmed.
Teams report 6 to 11% email replies with a 1.5 to 2x lift over email alone, when the email arrives three days after the request and extends the reference rather than repeating it. A bounced email after a request is a visible failure, so verify first.
Find the address, confirm it, and send the email three days later to a mailbox that exists. 100 free checks a month, unknowns refunded.
Ayoub built BounceZero's 5-stage validation pipeline, its dedicated BGP-announced IP infrastructure, and the Patroni HA PostgreSQL cluster behind every verification. Previously built high-volume email delivery infrastructure. Trained at 1337 Benguerir (École 42 network, 2019). Open-source: bgp_analyzer.
Templates by trigger and stage, each with the deliverability rules that apply
A recruiter outreach email template for contacting candidates cold
A SaaS trial nudge email template for signups who stopped
Four SDR cold email templates under 80 words each
A consultant cold email template for selling judgement rather than a product
A financial services cold email template built for compliance
A real estate agent cold email template for prospecting owners, landlords and local businesses
Continue through related topics