---
title: "Cold DM Opener That Gets a Reply: 14 Scripts"
slug: cold-dm-opener-that-gets-a-reply
author: "Leonardo Maldonado"
category: "DM Closing & Scripts"
articleType: listicle
tags: ["cold DM templates","DM opener that gets replies","how to start a cold DM","cold DM scripts for coaches","cold DM message request"]
publishedAt: 2026-09-29T09:00:00.000Z
updatedAt: 2026-09-29T09:00:00.000Z
canonical: https://setluca.com/blog/cold-dm-opener-that-gets-a-reply
---# Cold DM Opener That Gets a Reply: 14 Scripts

> A cold DM opener that gets a reply is short, specific, and honest about being cold. Match it to how cold the lead really is: mutual connection, content engagement, event or community, referral, or no connection at all. The 14 named openers below are copy-paste ready. Luca does not send them. Luca is reply-only, and it starts work after the lead answers.

## Key takeaways

- The best cold DM openers are specific, short, and ask for a low-effort reply with a clean exit.
- Permission-based framing outperformed interruptive framing in Gong's 300M-call study, 11.18% against 2.15%.
- Sort leads by how cold they actually are. A mutual connection and a hashtag stranger need different first lines.
- On Instagram, a cold message lands in the requests folder, and you get one message until they answer.
- Luca is reply-only: it drafts and sends your replies automatically in your voice once a lead responds, and you can switch any channel to review if you'd rather check first.

---

A cold DM opener that gets a reply is short, names something true and checkable about the person, and is honest about being cold. Cold is not one thing, though. A stranger who's watched your stories for a month is barely cold. A handle you found in a hashtag search is properly cold, and that message has to carry more honesty and less confidence. Below are 14 named openers sorted by that gap, plus the platform rules that decide whether your message is even delivered.

## Why do most cold DM openers get ignored?

Most cold DM openers get ignored because they lead with the sender. The inbox is where people talk to their sister and their group chat, so a message that reads like a broadcast is obvious from the first four words, and gets treated the way broadcasts get treated.

Coaches also rush cold outreach because they quietly hate doing it. Nobody enjoys writing to a stranger who didn't ask. So the research gets skipped, the line gets generic, and the message that goes out is the one least likely to work.

## How cold is your lead, really?

Sort by evidence, not by feeling. The question is what they already know about you, and what you can prove you know about them. Permission framing is the shape with hard numbers behind it: permission-based openers booked meetings 11.18% of the time against 2.15% for the interrupting "did I catch you at a bad time?" line, across more than 300 million cold calls (Gong Labs, 2024).

**Chart:** In Gong's 300M-call study, permission-based openers succeeded 11.18% of the time versus 2.15% for the interruptive 'did I catch you at a bad time?' opener. (Source: Gong Labs, 2024 (300M+ cold calls).)

The same effect shows up inside a single sentence. Across 90,380 recorded first calls, "How have you been?" hit 10.01% against a 1.5% baseline, while "Did I catch you at a bad time?" cut booking rates by 40% (Gong Labs). Those are calls, not DMs, and the part that travels is the mechanism: one low-pressure sentence decides whether a stranger leans in or looks for the exit. For inbox-native numbers, [DM opener performance data](/blog/dm-opener-performance-data) breaks reply rates down by opener type.

| How they got on your list | What you can honestly claim | Openers to reach for | Shape to use |
| --- | --- | --- | --- |
| Mutual connection or shared group | Membership, a checkable name | The Shared Room, The Common Client, The Second-Degree Ask | Personalized |
| Engaged with your content | Their own repeat behaviour | The Quiet Watcher, The Save, The Third Like | Question |
| Event or community | A shared room, a short window | The Room We Were In, The Session Note, The Public Answer | Specific-observation |
| Referred by a client or peer | Borrowed trust | The Warm Handoff, The Cold Referral | Personalized |
| No connection at all | Only what's public | The Flag, The One Observation, The Standing Offer | Permission-based |

Use a mutual-connection or referral opener whenever a real name can be dropped honestly, because a checkable name beats any clever line. Use a content-engagement opener when their own behaviour is the evidence, and keep it light so it doesn't read as surveillance. Use a cold opener only when someone's public work makes them an obvious fit, and send those by hand in small numbers. If no row fits, go and earn one before you message.

## 14 cold DM openers, by context

Swap the brackets for real detail. If you can't fill one with something true about that person, you've picked the wrong opener. All 14 are meant to be sent by hand.

### When you have a mutual connection

**1. The Shared Room**

> "Hey [name], think we're both in [group]. Saw your post about [specific thing] and it stuck with me. Still working on that, or did you sort it?"

_Use when_ you can name a group you're genuinely both in. _Why it works:_ the room supplies the context a cold DM lacks, so you're a member they haven't met rather than a stranger. _If they push back:_ "Fair, I don't message people out of nowhere often." Then let it go.

**2. The Common Client**

> "Hey [name], we've both worked with [mutual name]. I coach [niche] on [problem], and your last post read like you're in the middle of it. Am I reading that right?"

_Use when_ a shared client would vouch for you. _Why it works:_ a checkable name moves you out of the scam bucket in one line. _If they push back:_ offer the check. "Ask her, she'll confirm it."

**3. The Second-Degree Ask**

> "Hey [name], [mutual] and I talk about [topic] a lot and your name came up. Not pitching you. Curious what you're doing about [problem] at the moment."

_Use when_ the mutual hasn't offered an introduction, so you can't claim one. _Why it works:_ "not pitching" is only believable when the next sentence is a question. _If they push back:_ believe them and leave the door open once.

### When they've engaged with your content

**4. The Quiet Watcher**

> "Hey [name], you've caught most of my stories on [topic] lately, so I'll assume it's relevant. Is [problem] something you're dealing with, or are you watching for a client?"

_Use when_ they watch your stories consistently and have never messaged. _Why it works:_ their repeat behaviour is the personalization, so you don't invent a compliment. _If they push back:_ "Fair enough, Instagram just shows me who's watching."

**5. The Save**

> "Hey [name], you saved the [topic] post. Want the longer version of any of it?"

_Use when_ they saved or shared a post and the app showed you. _Why it works:_ a save is intent without commitment, and you're offering what they already reached for. _If they push back:_ send it anyway with no ask attached.

**6. The Third Like**

> "Hey [name], you've liked the last few [topic] posts, so quick question: is [problem] the thing you're working on now, or just interesting?"

_Use when_ there are repeated likes, no comments, no DMs. _Why it works:_ "just interesting" is a real answer they're allowed to give, which is what makes the yes worth something. _If they push back:_ nothing. A like isn't permission.

### When you met at an event or share a community

**7. The Room We Were In**

> "Hey [name], we were both at [event] in [month]. Your question in the [session] about [topic] was the best one asked. Did you get an answer that actually worked?"

_Use when_ it's within two weeks of the event and you can name what they said. _Why it works:_ events open a short window where a stranger's message is expected. _If they push back:_ "No worries, it was a good room."

**8. The Session Note**

> "Hey [name], we were on the same [group call]. I wrote down that you're [detail from their intro], which is the exact thing I work on. Worth swapping notes, or are you sorted?"

_Use when_ people introduced themselves on a small-group call. _Why it works:_ quoting their intro proves attendance, and "or are you sorted?" hands them a clean no. _If they push back:_ take the no and ask nothing else.

**9. The Public Answer**

> "Hey [name], saw your question in [community] about [problem]. I answered in the thread, but the honest version is longer than a comment. Want it here?"

_Use when_ you've already answered them publicly, for free. _Why it works:_ you gave before you asked, and they can go and check that you did. _If they push back:_ post the long answer publicly instead.

### When someone referred them

**10. The Warm Handoff**

> "Hey [name], [referrer] said she'd mention I'd be messaging. She told me about [problem]. What's actually going on with it?"

_Use when_ the referrer has told them you're coming. _Why it works:_ you're expected, so the message goes straight to their situation instead of your credentials. _If they push back:_ "She may not have got round to telling you."

**11. The Cold Referral**

> "Hey [name], [referrer] mentioned you when I asked who's serious about [topic]. He doesn't know I'm messaging, so feel free to ignore this. Still stuck on [problem]?"

_Use when_ the name is real but the introduction isn't. _Why it works:_ admitting the referrer didn't send you is what separates this from a name-drop scam. _If they push back:_ "Understood, I'll leave it there."

### When it's genuinely cold

**12. The Flag**

> "Hey [name], this is a cold DM and I'll keep it to four lines. I coach [niche] on [problem]. Your [specific post] made me think you're mid-[situation]. Want the two things I'd change, or shall I leave you alone?"

_Use when_ there's no connection and their public work makes them an obvious fit. _Why it works:_ naming the message defuses it, and the exit is what makes a yes mean something. _If they push back:_ take the exit you offered.

**13. The One Observation**

> "Hey [name], one observation, then I'll go. [Specific, checkable thing about their offer, funnel, or content.] Deliberate, or a gap?"

_Use when_ you can point at something concrete on their profile. _Why it works:_ a checkable observation proves you looked, and the question is answerable in one word. _If they push back:_ "Deliberate is a good answer." Then stop.

**14. The Standing Offer**

> "Hey [name], not asking for a call. I put together [specific resource] for [niche] dealing with [problem]. Want me to send it, or no?"

_Use when_ you have something genuinely useful and no relationship. _Why it works:_ yes-or-no to a file is the smallest ask in sales, and "or no" is doing half the work. _If they push back:_ send it, then say nothing for a week.

For a fuller library sorted by inbound entry point instead of coldness, see [Instagram DM openers that get replies](/blog/instagram-dm-openers-that-get-replies).

## What happens to a cold DM in Instagram's message requests?

Before you judge an opener, check whether it was ever delivered. A DM to someone who doesn't follow you lands in their message requests, and plenty sit unopened for months. Your reply rate measures two things at once: whether the line worked, and whether they ever looked.

The practical rules are tight. You get one message into a request thread. Until they accept or reply you can't send a second, and you can't attach media or voice notes the way you can in an accepted thread. That single message is the whole campaign, which is why a four-line message with an easy question beats a two-word "Hey."

Automation narrows it further. Meta's Messenger Platform and Instagram Messaging API policy gives businesses a 24-hour window to respond after a person messages first, and that window is what grants permission. Outside it, unsolicited messaging isn't allowed for standard use cases, and people can block or mute a business at any time. No API path exists for a compliant cold first DM, so if a tool offers you one, the account carrying the risk is yours. See [what Meta allows for DM automation](/blog/what-meta-allows-for-dm-automation) and [how many DMs you can send on Instagram per day](/blog/how-many-dms-can-you-send-on-instagram-per-day).

## Should you personalize deeply or send at scale?

Deep personalization wins, and the maths is less generous to volume than it looks. Say ten researched messages get a 20% reply rate: that's two conversations. A hundred blasted messages at 1% would get you one, plus reports and a throttled account. Cold DMs punish scale in a way cold email doesn't, because the recipient can end your reach with two taps.

Depth also means something narrower than flattery. The useful kind is checkable: a post they made, a group you share, a gap anyone could verify. "Love your content" costs nothing to write and proves nothing, which is exactly how it reads on the other end.

| Approach | Messages per day | Research per message | What it costs | Use it when |
| --- | --- | --- | --- | --- |
| Deep and slow | 5-10, by hand | 3-5 minutes | Your time | Offer is $2k+, or you're testing new copy |
| Medium | 15-25, by hand | 60-90 seconds | Some reply-rate drop | You have a verified signal that predicts fit |
| Blast | 50+, templated | None | Reports, action blocks, reach | Never, on this channel |

Choose deep and slow if your offer is high-ticket or your account is new. Choose medium once you've found a signal like a hashtag or a competitor's comment section. If you want the third row, change the channel rather than the opener. [Is cold DMing dead](/blog/is-cold-dming-dead) covers where this one still pays.

## What makes a cold DM opener fail?

A warm lead who dislikes your message ignores it. A cold recipient reports it, and Meta's Community Standards carry a Spam policy covering unsolicited bulk messaging, which is what one line copy-pasted to 200 people looks like from the platform's side.

| What you sent | Why it fails |
| --- | --- |
| Fake familiarity ("Hey! How've you been?") to a stranger | They check the thread, find nothing, and read you as a scam. |
| "Quick question" with no question in it | Bait with no payload. They know the next message is a pitch. |
| A voice note as message one | Asks 40 seconds of attention from someone who owes you none. |
| A link in the first message | Requests folders and spam filters both treat unsolicited links badly. |
| Three paragraphs of your credentials | Nobody reads a CV from a stranger. Credentials land after they name a problem. |
| "I noticed you're not doing X" | Opens with an insult, closes with a pitch. Even when true, it stings. |
| A visible merge tag ("Hey {first_name}") | Proves it was a blast, in one glance, permanently. |
| A second message before they've replied | Often not even possible, and where it is, it reads as pressure. |
| Anything sent to a scraped or bought list | No signal, no fit, and the reports arrive fast. |

## Worked example: Nadia and the 40 cold DMs

Nadia coaches strength athletes coming back from injury. She'd been sending roughly 40 cold DMs a week off a hashtag search, all with the same line: "Hey! Love your training content 🙌 Do you work with a coach?" Over six weeks she got 11 replies, two useful, and one 24-hour action block.

She rebuilt it around evidence. Instead of a hashtag she pulled people who'd commented on rehab posts under three accounts in her niche, which gave her a checkable observation for every name. That took her from 40 names to 12. Four shared a Facebook group and got The Shared Room. Five had commented on a question she'd answered publicly and got The Public Answer. Three had nothing and got The Flag.

Here's the one she sent a lifter called Tomás:

> "Hey Tomás, saw your question in the rehab thread about pressing around a cranky shoulder. I answered in the comments, but the honest version is longer than a comment. Want it here?"

Tomás replied in about two hours: "yes please, physio cleared me but pressing still hurts." Nadia sent three paragraphs and no offer. Two days later he asked what she charged. That week she sent 12 messages, got 7 replies, booked 2 calls, and triggered no blocks. The part worth copying is that she made herself earn a checkable sentence before anyone got a message. _(Composite example, illustrative.)_

## How should you follow up on a cold DM?

Cold follow-up is a different discipline. With a warm lead who replied and went quiet, a multi-touch cadence is expected. With a cold lead who never replied, every extra message moves you closer to a report.

**One message, then stop.** If they never answered, you don't have a conversation to follow up on. On Instagram the request rules usually enforce this for you.

**If they accepted but didn't reply, you get one more.** Wait five to seven days, make it shorter than the first, and hand them the exit: "No worries if it's a no, just didn't want it sitting there unanswered."

**If they replied and then went quiet, treat them as warm.** They opted in by answering, so a real sequence is fair. [Our DM follow-up sequence](/blog/dm-follow-up-sequence) has the timing.

**Never re-open a dead cold thread with new copy.** Coming back three months later with a different pitch to someone who ignored you twice is the behaviour the spam policy exists to stop.

## Which edge cases does most cold-DM advice skip?

**They're a peer, not a prospect.** Plenty of the people who look like ideal clients sell what you sell. Check the bio for a services link first. If you get it wrong, say so plainly and talk shop instead. Peers send referrals.

**They accept the request and say nothing.** An accept without a reply is a soft maybe. Wait a week, send one short line, then leave it. Accepting isn't consent to a sequence.

**They tell you to stop.** Reply once with "understood, sorry to bother you" and never message that account again. Not from a second account, not next quarter.

**They already told you no, months ago.** A previous no makes the next message worse, because now you look like a system rather than a person. Only re-open if something changed and you can name it.

**Their account is private or nearly empty.** There's nothing to personalize with, so either skip them or send The Flag with no fake specifics in it. A hollow compliment is worse than no message.

## Why isn't your cold DM opener getting replies?

| Symptom | Likely cause | Fix |
| --- | --- | --- |
| Sent but never opened | Sitting unopened in the requests folder | Get visible first. Comment on their posts for a week so your name is familiar when the request lands. |
| Action block after a batch | Too many non-follower messages in a short window | Stop for 48 hours, then resume at a third of the volume. |
| Replies are mostly "who is this?" | Your profile does no work for you | Fix the bio, pin three posts that show the problem you solve, then resume. The profile is part of the opener. |
| Opened, read, never answered | The question needs a paragraph to answer | Make it a two-option question. "Self or client?" beats "tell me about your training." |
| Reply rate collapsed in week three | One line at volume, now pattern-matched | Rotate two openers per context and cut daily sends until it recovers. |
| Getting blocked or reported | You pitched, linked, or followed up before they replied | Remove the ask from message one entirely for a month. |
| Good replies, no calls booked | Message one is fine, messages three to eight aren't | Audit the middle. Instagram DM scripts to book calls has the ask. |

## Which mistakes do coaches make with cold DMs?

**Treating cold like warm.** Warm openers assume a shared context the recipient doesn't have. Sent cold, "so what did you end up deciding?" reads as a scam.

**Scaling before the line works.** It's tempting to send 200 messages to learn whether an opener works. Twenty tells you most of what you need and doesn't get you throttled.

**Automating the first message.** No compliant API path exists for a cold first DM, so every tool offering one is working around the platform. The restricted account is the coach's, not the vendor's.

**Assuming a follow is permission.** Following someone changes where your message lands, not whether it was invited.

**Optimising copy when the list is the problem.** If these people can't afford or don't want your offer, no first line fixes it. Reply rate is a list metric before it's a copy metric.

## What happens when cold DMing goes wrong?

Sending a genuine cold message by hand keeps you outside the API rules, but not outside the Community Standards. Meta's Spam policy covers unsolicited bulk messaging, and its messaging policy makes the 24-hour window after someone contacts you the basis for business messaging. Once they message first, a [Facebook Messenger bot for lead qualification](/blog/facebook-messenger-bot-for-lead-qualification) can work inside that window.

The failure ladder runs in order. People mute or ignore first, which shows up as an unexplained drop in reply rate. Then they block, which quietly removes you from their view. Then they report, and enough reports in a window trigger an action block. After repeated blocks come broader account restrictions, and reach usually drops before anyone tells you why.

Recovery is unglamorous. Stop all cold sending for at least 72 hours. Don't switch to a second account, because linked accounts get caught together. Resume at a fraction of your old volume, hand-typed only, and rebuild normal activity: replies, comments, stories. [How to fix an action block on Instagram](/blog/how-to-fix-an-action-block-on-instagram) covers the steps.

One more thing worth saying out loud. If anything in your DM flow is drafted by AI, don't deny it when a lead asks. A plain "yes, I use an assistant to help with replies" costs nothing and keeps the thread honest.

## The honest limit: openers start conversations, and Luca doesn't send them

Openers only start the conversation. What you send next decides whether it converts, and cold threads tend to die in messages three to eight, when the replies turn slow, vague, or pushy. Learn to close without pressure in our guide on [how to close clients in the DMs](/blog/how-to-close-clients-in-dms).

There's a second limit, and this post is built around it. Luca is the DM sales platform for coaches, creators, and experts; it writes replies in your voice across Instagram, WhatsApp, Telegram, and Messenger, and it's reply-only by design. Luca never sends a cold first message. If cold outreach at volume is your strategy, Luca is the wrong tool, and we'd rather say so than sell you something that doesn't do it.

What Luca does is everything after the reply. Once someone answers, Luca drafts a personalized response that matches your tone and references what they said. Every reply sends automatically in your voice by default, and you can switch any channel to review if you'd rather check messages first, which is worth doing in an early conversation that's easy to misread.

Software also can't feel the room the way you can when a cold thread turns personal. A stranger answering your opener with a real problem is the moment to take over yourself. Human judgment stays on the first line and the close, while Luca handles the speed and the follow-up. See plans and limits on [Luca's pricing](/pricing), and learn the softer approach in our guide to [selling in the DMs without being salesy](/blog/how-to-sell-in-the-dms-without-being-salesy).


## FAQ

### What makes a cold DM opener get a reply?

A cold DM opener gets a reply when it's specific, short, and easy to answer. Reference something real about the person, keep it to a sentence or two, and ask for almost nothing. Gong's analysis of 90,380 first calls found the opening line alone swung success rates from 0.9% to 10.01%, so the sentence is worth the extra minute.

### How is a cold DM opener different from a warm one?

A warm opener can assume shared context. A cold one has to supply it. Cold openers name where you came from, admit the message is unexpected, and offer an easy exit. On Instagram a cold message also lands in requests, so you get one message until they answer, which means no follow-up to rescue a weak first line.

### Should I personalize every cold DM opener?

Yes, and the personalization has to be checkable. A specific first line proves you looked and separates you from a blast. McKinsey's 2021 Next in Personalization report found 71% of consumers expect personalized interactions and 76% get frustrated without them. "Love your content" isn't personalization, because anyone could have written it.

### Are permission-based openers better than direct pitches?

Permission-based framing outperforms interruptive framing. In its 2024 analysis of 300M+ cold calls, Gong Labs found permission-based openers succeeded 11.18% of the time versus 2.15% for "did I catch you at a bad time?" The framing lowers resistance. Acknowledge the message is unexpected, then ask for a small amount of attention.

### How many cold DMs can I safely send a day?

Fewer than you want to. Instagram throttles accounts messaging many non-followers quickly, and new accounts get the tightest limits. Five to ten hand-written messages a day is a sane starting point for a coach, and the honest constraint is research time: if you can't write a checkable sentence about someone, they shouldn't get a message.

### Can I follow up on a cold DM that got no reply?

Usually not, and usually you shouldn't. On Instagram the request-folder rules block a second message until they respond. If they accepted the request but stayed silent, one short follow-up after five to seven days is fair. If they never accepted, treat the silence as an answer.

### Is cold DMing against Instagram's rules?

Sending a genuine message by hand isn't banned, but volume cold DMing runs into Meta's Community Standards Spam policy on unsolicited bulk messaging. Automated cold outreach is a clearer problem: Meta's messaging policy grants a 24-hour window only after a person contacts you first, so no compliant API path exists for a cold first message.

### Is Luca a cold outreach tool?

No. Luca is reply-only and never sends a cold first message on any plan or channel. It works once a lead replies, drafting and sending personalized follow-ups automatically in your voice, with review available per channel if you want to check them first. RAIN Group's survey of 488 B2B buyers found 71% want to talk to sellers with new ideas, and that first, opted-in step stays yours.


## Sources

1. [Gong Labs -- The Best Cold Call Openers Backed by 300M Calls](https://www.gong.io/blog/the-best-and-worst-cold-call-openers-backed-by-data-from-300m-calls)
2. [Gong Labs -- Effective Cold Call Opening Lines (90,380 calls analysed)](https://www.gong.io/blog/cold-call-opening-lines)
3. [RAIN Group -- When Do Buyers Want to Talk to Sellers? (488 B2B buyers)](https://www.rainsalestraining.com/blog/when-do-buyers-want-to-talk-to-sellers-the-time-is-now)
4. [McKinsey -- The Value of Getting Personalization Right (Next in Personalization 2021)](https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights/the-value-of-getting-personalization-right-or-wrong-is-multiplying)
5. [Meta for Developers -- Messenger Platform and Instagram Messaging API Policy Overview](https://developers.facebook.com/docs/messenger-platform/policy/policy-overview)
6. [Meta Transparency Center -- Community Standards: Spam](https://transparency.meta.com/policies/community-standards/spam/)

---

Published by SetLuca, the company behind Luca.