I compiled a list of 50 LinkedIn DM's that worked for me

LinkedIn + Reddit at 90% of customers is a strong foundation ,but both are pull channels. You're either catching people in the act of searching or catching them mid-intent on LinkedIn. Curious whether you've tested any push channels that create demand upstream, before people are actively looking for a lead gen tool.

180k DMs and tracking only the winners — that's disciplined. The pattern I'd love to see: what's the one opener that worked across EVERY industry? Building Bexra (Helping entrepreneurs find, build & grow) and cold outreach is our next frontier. Also curious — did you find any difference in reply rates between InMail vs connection requests with notes?

Great list. One thing I'd add - personalization at scale is the real unlock. We've been testing AI-generated video in cold outreach and the reply rates are significantly higher than text-only

Thank you so much. I’ll definitely keep that in mind and apply it.

The "no pitch in message 1" rule is the one that's hardest to actually follow; you know your product is relevant, you know the hook is good, and the temptation to just drop the whole thing in one message is enormous. But every time I've tested short opener vs full pitch, the short one gets more replies. The discipline of saving the pitch for message 2 feels like you're leaving something on the table, but it's the opposite.

Thanks, this is extremely useful

The swipe file is useful, but the bigger lever for B2B outbound is which intent trigger you're paired with, not which opener. Looking at IbexAI's positioning, it sounds like the product itself is meant to bridge that gap by finding the trigger first, then handing off the message. From running outbound through an MSP for years, our highest reply rates were when the opener referenced a specific event the prospect had just experienced (funding round, tech migration, role change), not when the opener was clever in isolation. Curious if you have data from those 50 winners on how many were paired with a real-time intent signal vs cold-targeted to a list.

The meta-lesson here is underrated: you ran 180k DMs but you were also tracking outcomes at every step. Most founders skip that second part entirely.

The same pattern shows up in product analytics — people instrument everything, fire all the events, build the dashboard... but never ask "what does a successful outcome actually look like in the data?" So the pipeline runs, the metrics look clean, and nobody realizes they've been measuring the wrong thing for months.

What you're describing — logging what worked and finding patterns — is just disciplined data work applied to outreach. The rare part isn't having a swipe file, it's having had the discipline to log it while running live.

Similar reason I help early-stage founders audit their product data before scaling. A lot of "confident" growth metrics fall apart under one layer of SQL. Put together some free diagnostic scripts for exactly this → https://growthwithshehroz.gumroad.com/l/psmqnx

The 55% reply rate stat is wild by the way. What industry vertical were those outlier DMs hitting?

The per-industry breakdown is the part I would actually use. One pattern I'd be curious about: across the 50 openers that got real replies, was there a recognizable archetype that crossed industries, or did the winning approaches diverge sharply by buyer (SaaS founders responding to a fundamentally different opener than HR ops, say)?

The other question is more about methodology than the file itself. When you classified a reply as 'engaged', what was your threshold? I have run sequences where a 'real reply' rate looked great on the surface but the conversion to a meaningful conversation was much lower. The opener that maximizes reply rate often loses to one that filters harder upfront, even though the latter looks worse on paper.

Curious if any of the 50 in your file felt like that. Modest reply rate, but the conversations that followed were disproportionately useful.

One thing we've been noticing as well is that most outreach failure is not actually a messaging problem — it's a follow-up consistency problem.

A lot of founders optimize the first DM endlessly, but the real drop-off happens after the first touch. The lead replies mentally before they reply operationally, and then the conversation disappears because nobody owns the follow-up cadence.

The interesting part is that relevance and timing matter more now than sheer volume. Short, context-aware outreach combined with disciplined follow-up seems to outperform aggressive automation almost everywhere.

Really solid dataset here. Logging positive-response openers instead of just “best practices” was the smart move.

I’ve tested a lot of outreach angles for SEO/blogging services too, and the biggest lesson was this: the opener matters way more than the pitch. Most people lose the reply before the conversation even starts. Love that you actually tracked patterns instead of guessing.

The pattern I've noticed too: the DMs that work lead with curiosity about their business, not a pitch about yours. "How are you currently handling X?" beats "I built a tool for X" almost every time. Nobody responds to a solution before they've confirmed the problem. Bookmarking this list — the SaaS and recruiting/HR tech sections are exactly what I needed.

This aligns with something I’ve been learning while running ads for my marketplace.

I originally thought my biggest problem was onboarding because users weren’t creating listings after signup.

But after digging into the search terms and traffic behavior, I realized a lot of the issue was intent mismatch upstream.

People searching broad terms like “make money online” behaved completely differently from people already showing intent to offer local services or rent equipment.

The interesting part is that once intent quality improved, even small engagement signals became more meaningful.

Your point about warm signals vs cold interruption is huge. Timing and relevance seem to outperform “perfect copy” almost every time.

The swipe file approach is the right way to learn this, tracking what actually got a reply rather than theorizing about what should work. Went through the Notion list and the SaaS openers in particular are the ones that land differently. The ones that reference something specific about the person's situation outperform the ones trying to be clever almost every time. The pattern that jumped out: the best openers read like the start of a conversation, not the start of a pitch. I've been applying the same logic to IH threads ahead of my Product Hunt launch on May 13th, warm relationships convert at a completely different rate than cold clicks. What's the opener category you'd say works least well despite people using it most?

Thats an great approach

This is gold. Most people talk about “personalization” but rarely show actual examples that worked. Interesting that you tracked only the first message — did you notice any patterns across industries that consistently performed better?

THat is so helpful

Make sense, I'll surely strive in applying this.

This is a masterclass in turning distribution data into product insight. Most outreach tools focus on volume or personalization tokens, but you identified the real problem: cold outreach has a conversion ceiling because you're interrupting intent, not capturing it.

The 180K DM dataset revealed something critical: the winning openers weren't clever copywriting, they were timing alignment. When someone already signaled interest (commented on a relevant post, engaged with industry content), the opener almost didn't matter.

IbexAI essentially automates that signal detection, turning cold outreach into warm follow-up. The Harvard backing adds credibility, but the real validation is in the swipe file: 50 openers that worked aren't just templates, they're proof points that intent matters more than message craft.

One tactical question: are you segmenting signals by recency? A comment from 48 hours ago vs. 2 weeks ago likely has very different conversion rates.