LinkedIn Reply Rate Benchmarks by Industry: A Framework for Reading Yours

Why "the average LinkedIn reply rate" is the wrong number to chase

Every LinkedIn benchmark post leads with one headline number, something like "the average reply rate is 10%," and every reader mentally compares their own campaign against it regardless of what they sell or who they sell to. That single number is real, but it flattens a spread wide enough to make a genuinely good campaign look broken, and a genuinely broken one look fine.

I'm Hlib Storchak. I build outbound systems for B2B founders and sales teams, and I've booked 2000+ meetings for B2B clients running them. Reading a client's LinkedIn numbers against the right baseline, not a generic one, is one of the first things I do on any new account. The gap between industries in the data below is large enough that skipping this step is the single fastest way to make the wrong call on a campaign.

What Expandi's 2026 dataset actually measured

The most current and largest dataset I could verify directly is Expandi's "LinkedIn Outreach Benchmarks 2026" report, covering 13,218,869 connection requests, 6,730,447 outbound messages, and 3,766,161 accepted connections sent through 13,302 active accounts between May 2025 and April 2026. It tracks three separate numbers, not one, and treating them as interchangeable is the first mistake worth avoiding.

Acceptance rate also swings hard by industry on its own, from 17.5% in Consumer Electronics up to 40.1% in Broadcast Media, a point I've covered in more depth in what a normal LinkedIn acceptance rate actually looks like. This piece is about the reply-rate side specifically, where the industry spread is arguably more consequential, since a reply is what actually starts a conversation.

Staffing & Recruiting vs Computer Software: the widest gap in the data

Two industries sit at opposite ends of the reply-rate range, and both are common ICPs for B2B outbound teams, which makes the gap directly actionable rather than a trivia point.

Staffing & Recruiting runs roughly double the platform average on every one of the three metrics at once, not just one. Computer Software, meanwhile, is both the single largest industry by volume in the dataset and one of the weaker performers on reply rate specifically, even though its acceptance rate sits close to the platform average. A software sales team benchmarking its own 9% message reply rate against the flat 10.4% average would reasonably think it's underperforming slightly. Benchmarked against its actual industry peers, that same number is close to normal.

Why recruiting messages get replies and software pitches don't

The gap isn't really about copywriting quality, it's structural to what each message is offering. A recruiting outreach message is frequently something the recipient has a standing, ongoing interest in, a role, a client, a candidate, so the message itself carries reciprocal value independent of how well it's written. A cold sales pitch into Computer Software is competing with every other vendor DM a technical buyer already ignores by default, and the recipient has no standing reason to want the conversation before they've evaluated the offer.

That distinction matters beyond these two industries specifically. Any industry where LinkedIn outreach maps to something the recipient is already looking for, hiring, partnerships, referral-driven services, tends to sit above the platform average for the same reason. Any industry where outreach is a pure cold sales interruption, particularly one already saturated with vendor DMs, tends to sit below it.

Acceptance rate and reply rate are not the same problem

Computer Software's acceptance rate (27.5%) sits close to the platform average (28.5%), while its message reply rate (8.8%) sits well below the average (10.4%). That split tells you something specific: the profile and connection note are doing a reasonable job getting accepted, but the follow-up message isn't landing. If you only tracked one blended "engagement rate," you'd miss which half of the funnel actually needs the fix.

I've written the fuller breakdown of which metric to watch at which funnel stage in LinkedIn outreach metrics that actually matter. For this piece, the operative point is narrower: a low reply rate with a normal acceptance rate points at your message, your offer, or your timing, not at your profile or your targeting.

A 5-step framework for benchmarking your own numbers

Run these five steps before deciding a campaign is underperforming, rather than reacting to the first number that looks low.

1. Pull your own trailing 90 days, split by connection acceptance, note reply, and message reply. Treat these as three separate funnel stages, not one blended score.
2. Find your closest published industry analog. If your exact vertical isn't in a named dataset, use the closest adjacent one and adjust down slightly if your offer is a cold interruption rather than something the recipient is actively seeking.
3. Check whether the weak stage is acceptance or reply, not both. A weak acceptance rate points at profile and targeting. A weak reply rate with normal acceptance points at message and offer.
4. Segment by list quality before you blame the industry. A broad title-and-industry pull will underperform a tightly built list inside the same industry, so confirm the list itself is specific before concluding the industry is simply harder.
5. Re-check after a full sending cycle, not after a few dozen sends. A week of data is too noisy for a 10-20% baseline reply rate to read as signal rather than noise.

What to do when your industry isn't in the published table

Expandi's own report covers 60-plus industries, but the full breakdown table sits inside the report itself rather than being fully reproduced in every summary of it, so don't assume a specific number exists for your exact vertical just because the report claims broad coverage. When you can't find your own industry named directly, anchor on the nearest structural analog instead of the flat average: a services business selling into other services businesses behaves more like Staffing & Recruiting's reciprocal-interest pattern than like Computer Software's saturated-DM pattern, even if the industry label itself doesn't match.

The safer fallback, if you have any sending history at all, is your own trailing 90 days rather than any external number. An external benchmark tells you roughly where you should expect to land before you have your own data. Once you have three or four months of your own sends, your own baseline is a better read than someone else's industry average, since it already accounts for your specific list quality, offer, and sender profile.

What I check first when a client's LinkedIn numbers look bad

The mistake I see most often when a client tells me their LinkedIn numbers are "bad" is that they're comparing a single blended engagement number against a flat, industry-agnostic average, then deciding the whole channel isn't working. The first thing I actually check is the same three-stage split above: acceptance, note reply, and message reply, each against the closest industry read I can find, before touching a single word of copy. More often than the industry gap alone would suggest, the "bad" number turns out to be a completely normal reply rate for that specific industry, and the actual fix is elsewhere, usually the list or the offer, not the channel.

What chasing the wrong benchmark actually costs

This is an illustrative model built from stated assumptions, not a researched figure, so swap in your own numbers before treating it as a forecast. Assume a team running 500 LinkedIn connection requests a month into Computer Software, a genuine 9% message reply rate (in line with the industry data above), and a manager who benchmarks the team against the flat 10.4% platform average instead. Assume that misread costs one operator roughly a week of unnecessary campaign rewrites and a month of hesitation before scaling a channel that was already working.

At a fully loaded cost of roughly €4,500 to €5,500 a month for an SDR or outbound operator (a common range once on-costs are included), a week of misdirected rework is roughly €1,000 to €1,400 in wasted time, plus the opportunity cost of a month where a working channel got scaled back instead of scaled up. Run your own operator cost and campaign volume through the same logic: the real cost of benchmarking against the wrong number is rarely the analysis itself, it's the weeks spent fixing a channel that wasn't actually broken.

Three ways teams misread this data

These show up often enough to call out directly, independent of which industry you're in.

• Blending acceptance and reply into one score. They measure different parts of the funnel and point at different fixes, as the Computer Software example above shows directly.
• Benchmarking against the flat average instead of the closest industry read. A 9% reply rate reads as a problem against 10.4%, and reads as normal against Computer Software's own 8.8% baseline.
• Reacting to a few days of data. At typical B2B send volumes, a week's reply count is usually too small a sample to separate a real dip from ordinary week-to-week noise.

When industry isn't the real explanation

Industry benchmarking is a starting point, not a verdict. If your number sits well outside even your own industry's band, in either direction, the more likely explanations are list quality, sender profile completeness, message specificity, or send volume patterns that trigger platform-level throttling, not some property of your industry itself. Use the industry number to set expectations before a campaign starts, then diagnose against your own funnel stages once you have real data, rather than treating the industry average as the final word on whether a campaign is working.

FAQ