Annotated examples
Twelve real LinkedIn cold messages across SaaS, agency, fintech, dev-tools, consulting, and recruiting. Each one is followed by an annotation of which sentence does what, why the structure works, and which moves to copy into your own outreach.
A LinkedIn cold message example is a fully-written, real-world message reproduced as a teaching artifact — not a template with merge tags. The purpose of an example is to show how a working DM is constructed, line by line, so the reader can copy the structural moves into their own writing rather than copying the surface text.
Templates with {{first_name}} placeholders teach the form but not the craft. A real example shows the actual sentence rhythm, the actual word choice, the actual transitions — the things that separate a working message from one that reads like every other.
Reading a good DM teaches you the form. Reading a good DM with a sentence-by-sentence breakdown teaches you the craft. Annotations name what each line does so you can deploy the same move in your own writing.
A cold message that works for a SaaS founder selling to a CTO uses different moves than one from a recruiter to a passive candidate. The examples below span 7 distinct sender/receiver dynamics so you can find the one that maps to your own.
Sales reps and AEs at SaaS companies messaging buyers at their target accounts.
ScenarioSDR at a CRM hygiene tool, messaging a newly-hired RevOps lead.
Hey Maya — congrats on the RevOps role at Lumen. First 90 days at a Series B is usually equal parts "rip out the old reporting" and "figure out what the AEs will actually trust." We built a HubSpot ↔ Salesforce reconciliation layer that 3 RevOps leads I know shipped in their first month for exactly that. Happy to send the playbook + a 6-min Loom if it's relevant — no pitch. No worries either way, good luck with the ramp.
Line 1 — references a real signal (the hire) and a real moment (first 90 days). Line 2 — names a tension the reader genuinely feels ("what AEs will trust") in the language a RevOps lead would use themselves. Paragraph 2 — pivots to the product as "what peers shipped," replacing logo soup with operator credibility. Paragraph 3 — single-sentence sign-off with a no-follow-up signal that lowers the reader's guard.
ScenarioAE at a sales-coaching tool, messaging a VP Sales known to use a major competitor.
Hey Marcus — I know you're a Gong shop (saw your case study on their site, which I assume someone in marketing made you take). Not trying to displace it. We sit beside Gong and do one specific thing: turn the top 3 objections from this week's calls into a coaching micro-lesson your reps actually watch. Most teams that use both rate us as their highest-NPS sales tool. If you ever want a 90-second walkthrough on top of Gong, ping me. If not, no follow-up.
Line 1 — acknowledges the competitor and shows research with a wink ("which I assume someone in marketing made you take"). Paragraph 2 — explicitly disowns displacement, which removes the buyer's defensive crouch. The "do one specific thing" frame is honest positioning that respects the reader's intelligence. Final paragraph — calendar-link-free closer with a clear no-follow-up commitment.
Service businesses (agencies, consultancies, fractional executives) reaching potential clients.
ScenarioFounder of a B2B content agency messaging a head of marketing at a Series A SaaS company.
Hi Jordan — I run a small B2B content shop (8 people, work with Brex, Ramp, Pylon). Saw your job post for a content lead a few weeks back. Curious whether you ended up hiring, or if a fractional shop on a 3-month engagement would be useful. We usually replace the FTE hire for 4–6 months — let the team build proof points, then either hire a senior or extend us. About 40% of our clients do the latter. No Calendly link, no follow-up. If useful, just reply and I'll send 3 case studies.
Line 1 — names size, credibility (3 logos a marketing reader would recognize), and references a specific job post (proof of research). Paragraph 2 — frames the agency engagement against the FTE alternative, which is the buyer's actual mental model. Quoting the 40% extend rate is concrete in a way most agency cold messages aren't. Closer — no Calendly link, no follow-up promise, clear next step.
ScenarioFractional CFO reaching out to a founder whose company just raised.
Hey Sam — congrats on the Series A. I'm a fractional CFO who works with 6 Series-A SaaS companies right now (cap is 8). Most of them brought me in 4–8 weeks post-close, which is roughly when the board starts asking for monthly reporting that the existing finance setup can't produce. Not the right time for a full-time CFO at your stage, and most outsourced firms run too junior. I sit in the middle — senior, fractional, capped client load. Worth 25 minutes? I'll come with the financial model you should be running.
Line 1 — congrats + immediate credibility statement with a scarcity signal ("cap is 8"). Specifying "4–8 weeks post-close" shows pattern recognition that proves the sender has done this before. Paragraph 2 — positions against both alternatives (full-time CFO, junior outsourced firm) the way the buyer is actually evaluating. Closer commits to bringing real value to the call ("the financial model you should be running"), not just discovery questions.
Recruiters, sourcers, and hiring managers reaching candidates.
ScenarioIn-house recruiter at a Series B fintech, sourcing a senior engineer at a competitor.
Hey Ana — sourcing for the staff infra role at Stage. Tenure-wise you've been at Datadog 5+ years, which I'm guessing means you're either deeply happy or quietly ready for what's next. Either is fine — I'm not going to bug you twice. What's on the table: Stage is Series B, the infra team is 11 engineers, the work is rebuilding their multi-tenant database from Aurora onto Vitess. The kind of project most staff engs say they want and rarely get to ship. If you're curious enough to read a one-pager, send and I'll wrap up. If not, easy to ignore.
Line 1 — the "5+ years means happy OR ready" frame catches both groups without insulting either. Paragraph 2 — concrete technical pitch (stage, team size, project) in the language a senior engineer evaluates. Final paragraph — single-touch promise that respects the candidate's inbox. The recruiter is doing the rare work of being a normal human.
ScenarioEngineering manager messaging an engineer they want for their team.
Hey Tomi — I'm the eng manager hiring this role, not a recruiter, so I'll be brief. Your Strange Loop talk on feature stores was the best 25 minutes of that conference. I run the data platform team at Mosaic and we're about to start exactly that project. Looking for the principal eng who would lead it. If you're even mildly curious, 30 minutes with me to swap notes on the project? No Lever link, no take-home, just a conversation.
Line 1 — disclosing "I'm the manager, not a recruiter" is a credibility differentiator. Paragraph 2 — specific reference to their talk plus an exact match between the talk topic and the open project. Closer offers a low-friction conversation rather than a formal interview funnel, which is what passive senior candidates actually respond to.
Partnership and BD outreach where the ask is a relationship, not a deal.
ScenarioHead of partnerships at a customer-success tool, reaching a head of partnerships at a complementary in-product flow tool.
Hey Tessa — head of partnerships at Pylon here. We're seeing more and more of our customer-success customers asking for what Userflow does (in-product flows for new feature adoption), and I'm guessing your team sees the inverse for retention tooling. Worth a 25-minute chat about a referral motion? We've done this with 3 other adjacent tools — happy to share what worked, what didn't, and what a clean version would look like for us two. If yes, I'll send a couple of times. If now's not the right quarter, ping me Q1.
Line 1 — opens with an observation only someone in the role would make, positioning the sender as a peer. Paragraph 2 — quantifying "3 other adjacent tools" gives experience credibility and reduces partner risk. Final paragraph — gives the partner an honest off-ramp ("if now's not the right quarter") that often produces a "actually, now is fine" response.
ScenarioBD lead at a SaaS procurement tool, messaging a head of finance at a 500-person company.
Hi Renee — quick one. We work with 30 finance leaders at companies your size (200–800 employees) and the consistent story is: SaaS line items have grown 60–80% over 18 months while the actual number of seats per tool barely moved. We built a tool that finds the gap — typically $50–100k/year in unused or duplicate SaaS at a company your size. Worth a 20-min look? I'll bring our audit framework to the call so even if you don't buy, you've got something useful to run yourself.
Paragraph 1 — opens with a data point relevant to the reader's role, framed as a pattern across 30 peer companies. Specific size band ("200–800") shows targeting precision. Paragraph 2 — quantifies expected outcome with a dollar range, not a percentage. Closer commits to bringing real value regardless of buying decision ("audit framework even if you don't buy") — that promise differentiates this from every other BD message.
Technical founders and dev-tool sellers reaching engineering buyers.
ScenarioFounder of a dev tool messaging a staff engineer at a known customer of a competitor.
Hi Pat — I'm the founder of Bytebase. Saw you're running Liquibase at Vercel and have been for a while. Not trying to displace it — Liquibase is solid. But the use case where teams keep coming to us off Liquibase is: schema migrations across 20+ services where you need one approval workflow, not 20. If that's the shape of Vercel's problem, I'd send a 3-minute Loom. If not, you're using the right tool. Either way, follow you now — your post on the read-replica failover thing was sharp.
Line 1 — discloses founder identity. Paragraph 2 — explicitly compliments the competitor before pivoting, which is the move that distinguishes confident founders from desperate sellers. The "20+ services, one approval workflow" line is the kind of specific use-case description engineers respond to. Final beat — gracefully exits the pitch and ends on an unrelated genuine compliment, which closes the loop as a peer.
ScenarioMaintainer of an OSS project reaching a user at an enterprise to discuss enterprise support.
Hey Sam — maintainer of the Pgvector library. Github tells me your team at Stripe is one of our heavier users. We're starting an enterprise support tier — basically, a Slack channel with the core maintainers, plus a yearly contract for one critical-bug SLA. Mostly for teams that have shipped Pgvector to prod and would feel safer with a phone number. If that's a real need at Stripe's scale, happy to chat. If you're running it fine without, even happier.
Line 1 — credibility from being the actual maintainer, not a sales person at a company that wraps the OSS. Paragraph 2 — describes the offer in operator terms ("Slack channel with maintainers", "phone number") rather than enterprise-speak. Final paragraph — explicitly allows the prospect to opt out without consequence, which is the move that consistently wins respect from senior engineers.
Edge cases — investor outreach, reactivation, and a few moves you rarely see done well.
ScenarioPre-seed investor reaching a founder who hasn't announced raising yet.
Hey Aaron — early-stage investor at Heron. Saw the launch of OpenLayer last week and the design choice to keep the inference layer separate from the orchestration layer is, in my view, the only sensible architecture for what comes next. Most teams will look back in 18 months and wish they'd built it that way. Not a fundraising pitch — most founders at your stage aren't raising. But if you ever want a sounding board from an investor who has actually shipped this kind of system, that's an open offer.
Line 1 — opens with a specific technical observation that demonstrates the investor actually understands the product. Paragraph 2 — explicitly disowns the fundraising pitch, which is the single best move an investor can make in cold outreach. The "sounding board from someone who has shipped this" framing offers value upfront rather than asking for time.
ScenarioReaching a prospect who went dark on a previous evaluation 6+ months ago.
Hey Lila — we talked back in Q2 about pipeline forecasting tooling and the timing didn't land. Not relitigating any of it. Reason for the ping: we just shipped the integration with Looker (which I remember was the reason it didn't fit last time). Wanted to surface that in case the conversation's worth restarting. If not, all good — won't re-add you to anything. Either way, hope the year's gone well.
Line 1 — names the past conversation specifically (Q2, the topic, the reason it didn't fit) which makes the reach-out feel like a real continuation, not a templated reactivation. The reason for the ping is anchored on a real change that removes the original objection. Final beat — a human sign-off that lets the relationship close warmly if the answer is no.
The reasons the templates above work. Apply these and you can write your own without ever touching a template library again.
A template with merge tags teaches you the skeleton of a message. A real example with annotation teaches you the muscle. Read examples to learn the moves; use templates as the starting structure when you scale.
In every example above, the opener does the heavy lift — establishes credibility, references shared context, names a signal, or disowns a sales motion. If the first sentence is weak, no amount of strong middle paragraphs will save the message.
Half the examples above include an explicit disowning — "not trying to displace it," "not a fundraising pitch," "won't re-add you to anything." This counter-intuitively lifts trust because it shows the sender knows the bad version of the message and is choosing not to send it.
Multiple examples promise something the reader gets even if they say no — an audit framework, a financial model, a template. That commitment differentiates the message from every other cold outreach and removes the asymmetry of the ask.
The best closers are not calendar links. They are short human beats — "good luck with the ramp," "follow you now, keep the takes coming," "hope the year's gone well." These cost you nothing and make the message read as written by a person, not a sequence.
A template is a reusable skeleton with merge tags ({{first_name}}, {{company}}) designed to be sent at scale. A real example is a fully-written message reproduced as a teaching artifact — it shows the actual word choice, sentence rhythm, and structural moves a working message uses. Examples are more useful for learning; templates are more useful for scaling once you understand the craft.
Not effectively. The power of each example comes from a specific signal (a hire, a post, a competitor adoption) that is unique to one prospect. Copying the surface text will read as templated. Copy the structural moves — the opener pattern, the disowning sentence, the value-on-the-call promise — and rewrite the surface text for your own prospect.
Read each sentence and ask: what is this sentence doing? Establishing credibility? Referencing a signal? Pivoting to the pitch? Lowering the reader's guard? Closing as a human? When you can name what each sentence does, you can deploy the same move in your own writing. Annotation is a teaching shortcut that bypasses years of trial and error.
Well-crafted cold messages with strong personalization run 12–20% reply rates. Signal-triggered messages (job change, funding, competitor signal) often reach 25–35%. The examples above are calibrated to the upper end of those bands because each one is built around a strong reason-to-send. Generic mass-sent versions of the same templates would underperform dramatically.
Optimal length is 400–550 characters formatted as 2–3 short paragraphs. The examples above range from 415 to 559 characters, which is the band where the message has room for a real opener, a real pitch, and a real closer without crossing into "brochure" territory.
Both — for different tiers. For your top 50 target accounts, write each message manually using the structural moves from the examples above. For the long tail, use templates with AI personalization that adapts the opener to each prospect's recent activity. See /ai-personalization-for-linkedin for the working pattern in 2026.
Three tests. First, read it out loud — if you wouldn't say it to a peer at a coffee shop, rewrite. Second, ask: could the first sentence have been auto-generated? If yes, rewrite. Third, ask: would I feel embarrassed if this message were screenshotted on Twitter? If yes, rewrite. Messages that pass all three tests reliably outperform.
Connect a LinkedIn sender, paste any template above into a campaign, and launch with per-sender daily caps and AI personalization on every send. Free for 7 days. No credit card required.
