Voice notes on LinkedIn are worth testing, but they’re not a blanket upgrade over text: used to add warmth to a reply or a second-touch follow-up, they can noticeably lift response rates; used as a cold opener to a stranger, they tend to feel intrusive and get ignored or reported. The skill is knowing which stage of the conversation they belong in, not whether to use them at all.
TL;DR:
- Voice notes work best in warm or semi-warm moments — a reply to an accepted connection, a follow-up after engagement, or a thank-you after a call — not as a cold first message.
- Used as a cold opener, voice notes often feel like a bigger intrusion than text, because they demand the prospect’s time and attention immediately, with no easy way to skim.
- Keep them short: 20–40 seconds is the sweet spot. Anything over a minute starts to feel like a sales pitch rather than a message.
- Script a loose outline before recording — not a script read word for word — so the delivery still sounds human rather than rehearsed.
- Test voice notes against your existing text sequence on a like-for-like segment before rolling them out everywhere, and track reply rate and reply sentiment separately.
- Pair voice notes with a short text summary underneath for prospects who can’t or won’t listen (which is a meaningful share of any list).
Table of Contents
- What Are LinkedIn Voice Notes (and Why They’re Suddenly Everywhere)
- The Case For: Why Voice Notes Can Lift Reply Rates
- Where Voice Notes Backfire
- When to Use a Voice Note vs a Text Message
- How to Record a Voice Note Prospects Actually Listen To
- Scripting Your First Voice Note Message
- Measuring Voice Note Performance
- Common Mistakes to Avoid
- Building Voice Notes Into Your Sequence
What Are LinkedIn Voice Notes (and Why They’re Suddenly Everywhere)
LinkedIn added native voice messaging to its chat a few years back, but it has only really taken off in outreach circles over the last year or so, as reply rates on plain text sequences have kept sliding. The mechanic is simple: inside a LinkedIn conversation, you can hold down a microphone icon and record a short audio clip instead of typing. The recipient sees a waveform and a play button rather than a wall of text.
The appeal for teams running volume outreach is obvious. Text templates are easy to spot, skim, and ignore — prospects have seen a thousand variations of “great work at X” and tuned them out. A voice note is harder to fake at scale: it signals, rightly or wrongly, that a real person took thirty seconds to record something just for this prospect, which is valuable precisely because it’s still rare in B2B inboxes. It’s also worth being honest about why this matters now rather than two years ago: AI-written text has become good enough that pattern-recognition for “this was templated” has sharpened, and voice is one of the few channels AI tooling hasn’t fully colonised yet.
The Case For: Why Voice Notes Can Lift Reply Rates
Teams who test voice notes properly — on a controlled segment, against a text control, over real volume — tend to report a consistent pattern rather than a dramatic one. Reply rates on voice notes used in warm contexts (after a connection is accepted, after a prospect engages with a post, or as a follow-up rather than an opener) often run meaningfully higher than the equivalent text message, frequently 1.3 to 1.8 times the text reply rate, though this varies hugely by audience and seniority.
A few mechanisms likely drive that lift. Tone carries information text can’t: curiosity, confidence, or warmth comes through a voice clip in a way that’s hard to convey — and easy to get wrong — in two sentences of text. Voice notes also slip past the keyword-based skimming some senior prospects apply to their inbox, since the content isn’t visible until it’s played. And there’s a reciprocity effect: recording one takes visible effort, and prospects who listen through often feel a small pull to respond in kind — the same dynamic that makes a handwritten note outperform a printed one.
The lift is strongest with senior, time-poor buyers who are more likely to listen on the move than to sit down and parse a paragraph of text. It’s weakest with cautious, risk-averse personas — legal, compliance, public sector — who are less inclined to open unsolicited audio from a stranger.
Where Voice Notes Backfire
The failure mode is just as consistent, and it’s almost always the same mistake: using a voice note as the very first touch to a cold, unconnected prospect. In that context, a voice note from a stranger reads less like effort and more like an imposition — it demands 30 to 60 seconds of attention right now, with no way to skim the first line and decide whether it’s worth it, which is exactly what text lets a prospect do. Teams that test this find cold voice-note openers often underperform cold text openers on reply rate, and generate noticeably more “not interested” replies and blocks.
There’s also a practical access issue: a meaningful share of prospects open LinkedIn from a desktop at work, where playing audio without headphones is awkward, so a clip that can’t be conveniently played in the moment is often skipped and never returned to. And there’s a trust dimension specific to this moment: as AI voice cloning becomes more visible in the news, some prospects have grown wary of unsolicited audio from strangers, half-wondering if it’s even a real person — a suspicion that works directly against the “this feels more human” advantage voice notes are supposed to have.
When to Use a Voice Note vs a Text Message
The simplest working rule: use text to open, use voice to warm. In practice, that breaks down into a few concrete moments where a voice note tends to outperform text, and a few where text should stay the default.
Good moments for a voice note: immediately after a connection request is accepted (a brief, friendly “thanks for connecting,” not straight into a pitch); after a prospect likes, comments on, or shares a post, where you can reference the specific content; as a follow-up to a text message that got no response after 4–5 days, as a pattern-interrupt; after a discovery call, as a short recap-and-thank-you; and when re-engaging a dormant lead who went quiet months ago, where the novelty of audio can break through habituation to text.
Moments to stick with text: the very first message to a brand-new connection request; anything containing numbers, links, or scheduling details a prospect will want to reference later (nobody wants to replay a clip to catch a meeting time); outreach to personas known to be cautious about unsolicited audio (legal, procurement, public sector); and any touch where volume is too high to record a genuinely distinct clip per prospect — a voice note obviously reused for fifty people defeats the purpose and looks worse than a templated text message would.
How to Record a Voice Note Prospects Actually Listen To
Length is the single biggest lever. The sweet spot for most B2B contexts is 20 to 40 seconds — long enough to say something specific and personal, short enough that it doesn’t feel like a commitment to press play. Past 60 seconds, listeners disengage; past 90, you’re asking for a phone call’s worth of attention without the two-way benefit of one.
Environment matters more than people expect. Record somewhere quiet, ideally with a decent phone or headset mic rather than a laptop’s built-in mic picking up room echo and keyboard noise. A clip that sounds like it was recorded in a train station undercuts the “I took care over this” signal the format is meant to send.
Open with the person’s name and one specific, verifiable reference — a recent post, a comment, a company announcement — in the first five seconds. It proves the message isn’t mass-recorded and gives the listener a reason to keep going past the point where they’d normally bail. Close with a low-friction ask they can answer in one line of text, not “let’s book 30 minutes,” which is a much bigger commitment to extract from audio than from a text message with a calendar link attached.
Tone is worth rehearsing even if words aren’t. Recording two or three takes and listening back before sending is normal practice among teams who do this well — not to sound polished, but to catch the moments where nerves make a message sound flatter than intended. Smiling while you record, oddly, does change how a voice sounds, and prospects can tell.
Scripting Your First Voice Note Message
A full word-for-word script tends to produce a voice note that sounds like it’s being read, which defeats the purpose — but zero structure produces rambling, filler-heavy clips that lose the listener by the ten-second mark. The middle ground is a loose four-beat outline you internalise rather than read.
Beat one: name and a specific, personal hook that proves this clip is about them, not a template (“Hi Sarah, saw your post about the shift to in-house SDR teams — really resonated”). Beat two: the reason you’re reaching out, stated plainly, without a pitch (“wanted to flag something that’s come up a few times with firms in a similar spot”). Beat three: one concrete value point, not three (“we’ve seen teams in your space cut time-to-first-meeting by roughly a third just by changing how the first follow-up is sequenced”). Beat four: a low-friction close (“no pressure at all, just curious whether that’s something you’re seeing too”).
Practice the outline two or three times before recording for real prospects, so the beats become a rhythm rather than something you’re visibly reaching for mid-recording. The goal is a clip that sounds like you’d say the same thing if the prospect were standing in front of you, not a radio advert.
Measuring Voice Note Performance
Treat voice notes as a variant to be tested, not a blanket replacement. Run them against a genuinely comparable text control — same segment, same stage of sequence, same offer — and track play rate, reply rate, and reply sentiment separately, because a voice note can lift raw reply volume while also lifting the negative-reply share if it’s landing at the wrong stage of the relationship.
Most sequencing tools don’t natively track whether a voice note was played, so this often has to be tracked manually via a tagging convention for the first few hundred sends, until a pattern emerges you can trust. Expect a few weeks of testing before the data is reliable — reply rates on small samples swing wildly, and three good days can look like a trend that evaporates by week three. Segment your results by seniority and function where you can: it’s common to see voice notes dramatically outperform text with founders and senior commercial leaders, while barely moving the needle with operational or compliance-adjacent roles.
Common Mistakes to Avoid
The most damaging mistake is using a voice note as a cold opener to someone who’s never interacted with you, covered above — worth repeating because it’s the single easiest way to turn a promising channel into a source of blocks and spam reports. Close behind it is sending the identical audio file to dozens of prospects: the effort-signal that makes voice notes work is exactly what’s lost, and a prospect who recognises a reused clip reacts more negatively than to a reused text template, because the fakery feels more personal.
Other recurring mistakes: recording in a noisy environment that undercuts the message’s professionalism; going well over a minute and asking for a meeting’s worth of attention via a one-way channel; burying the ask so deep that a prospect who stops listening halfway never hears it; and skipping a short text caption summarising the gist for the significant minority of prospects who simply won’t play audio from a stranger.
A subtler mistake is treating voice notes as a volume play. They don’t scale the way templated text does — each one needs to be somewhat specific to land — so a sequence built to the same per-rep throughput as a text cadence usually ends in rushed, generic-sounding clips that get none of the benefit and all of the downside.
Building Voice Notes Into Your Sequence
In practice, the sequences that get the most out of voice notes treat them as a single, well-placed touch rather than a wholesale replacement for text — typically one voice note at the “thanks for connecting” moment or as the pattern-interrupt follow-up around day 5–7, with every other touch staying text. That keeps the recording load manageable and keeps the format feeling special rather than routine by the third or fourth clip a prospect receives from the same sender.
This is exactly the kind of channel-mix decision that benefits from having run enough sequences to know where a tactic earns its keep rather than guessing — which is where an agency partner doing done-for-you LinkedIn outreach, like The Lead Lab, tends to add value beyond just sending more messages: testing additions like voice notes against a real control group on real client data, and rolling them in only where the numbers support it rather than because the tactic is trending.
If you’re building this in-house, start small: pick one segment, one sequence position, and one rep willing to record consistently for two to three weeks. Compare the results honestly against the text control before expanding further, and resist rolling it out everywhere just because the early numbers look promising — early numbers on a new channel almost always look better than they turn out to be once novelty wears off.