The short answer.
In 2026 the platform average is roughly 28 to 30% connection acceptance and a 10.4% reply rate to messages after connecting. Good is 35%+ acceptance and 25%+ replies. You beat the benchmarks with relevance and timing, not volume: personalised, signal-timed messages written by a human reach 30 to 50% replies. That is exactly what Occura runs for you, on dedicated accounts, with in-house setters.
See how Occura would lift your reply rateWhat the 2026 benchmarks say.
A response rate means nothing without a baseline. The best dataset we have comes from analysing more than 13 million outreach attempts across 2025 and 2026, and it gives a clear picture of where the averages sit. Connection acceptance lands around 28 to 30%. The reply rate to a message sent after you connect holds at about 10.4%. The reply rate to the note attached to the connection request itself has slid to roughly 2 to 3%.
Those are averages, which means half of all campaigns do worse. What separates the good from the average is not luck, it is a small set of choices we will get to. First, the four numbers worth memorising.
For a working scorecard: below 20% acceptance or under 10% replies is weak, 25 to 35% acceptance and 10 to 25% replies is average, and the top performers clear 35 to 45% acceptance with 30 to 50% replies. The headline reply numbers also flatter you, so the metric that actually pays rent is the positive reply rate. Strong campaigns turn 5 to 8% of total sends into a positive reply, and book a meeting from 2 to 3% of sends.
A high reply rate full of “not interested” is a vanity number. Track positive replies and meetings booked, because those are the only two metrics that turn into pipeline.
Why reply rates are falling.
The averages hide a more important story: the direction of travel. Between May 2025 and April 2026, reply rates to connection-request notes fell by about 37%, from 3.5% down to 2.2%. Acceptance held steady and post-connection reply rates stayed flat, but that one channel collapsed.
The reason is not mysterious. The same dataset points to generic, AI-spun connection notes flooding inboxes, buyers becoming habituated to them, and LinkedIn quietly changing how it surfaces those notes. In other words, the more people automate templated openers, the worse templated openers perform for everyone. The volume tactic eats itself.
The reply rate did not fall because LinkedIn got worse. It fell because everyone started sending the same message. Relevance is now the only thing left that scales.
The strategic read is simple. As automated, generic outreach degrades, the gap between a templated message and a hand-written, well-timed one widens. The teams holding 30%+ reply rates in 2026 are the ones who never relied on the spray-and-pray approach in the first place. We cover the full picture of this shift in our complete guide to LinkedIn outreach.
Benchmarks vary by who you sell to.
There is no single “good” number, because your vertical sets the gravity. Recruiting and staffing sits near the top, with reply rates of 18 to 25%, partly because the message is genuinely welcome. Software and technology sits near the bottom, around 8 to 9% replies, because those inboxes are the most saturated outbound targets on the platform.
| Industry | Acceptance | Message reply |
|---|---|---|
| Civil engineering | 39.9% | 25.5% |
| Staffing & recruiting | 36.5% | 18.9% |
| VC & private equity | 34.9% | 11.0% |
| Computer software | 27.5% | 8.8% |
| Apparel & fashion | 19.9% | 7.1% |
Two findings cut against common assumptions. First, the sender's seniority barely moves acceptance: every title clusters within a three-point band, so a fancy job title in your outreach signature does not save a weak message. Second, company size does not matter either. A five-person startup gets the same acceptance rate as a ten-thousand-person enterprise. What moves the number is the message and the timing, full stop.
Comparing your software-startup campaign to a global 30% acceptance benchmark and panicking. Benchmark against your vertical, then beat it with the levers below.
The levers that actually beat them.
Beating the benchmark comes down to four levers, in order of impact. Each one is backed by the data, and none of them is “send more.”
1. A real message, not a template
This is the single biggest lever. After someone accepts, a personalised first message gets a 9.36% reply rate versus 5.44% with a generic one, and tailored messaging to a specific ideal customer profile lifts replies by up to 54.7%. The message is where the reply is won or lost, which is why we hand-write every one. Our connection request templates show the framework, but the variables have to be real.
2. Timing on a real signal
Reaching the right person in the wrong month looks identical to a bad list. Reach them inside the 14-day window after a trigger (a funding round, a new leader in the function you sell into, a relevant hire) and the same message converts several times higher. Building lists around live signals is the difference. Our guide to building a prospect list walks through how.
3. Follow-up, the part most people skip
It takes roughly five touchpoints to turn a prospect into a conversation, and multi-touch sequences spaced two to five business days apart convert about 49% better than a single message. Most replies live in the follow-ups, not the first send.
Connect
Personalised request tied to one specific reason.
Open
A short, relevant first message after they accept.
Give
Add a useful thought, not “just bumping this.”
Close the loop
A final, low-key check-in, then stop.
4. Brevity and a good send window
Short messages, under 400 characters, outperform longer ones by around 22%, and mid-week sends (Tuesday through Thursday, with Monday and Wednesday strongest) beat the weekend. These are smaller levers, but they are free.
Diagnose your own funnel.
Once you know the benchmarks, your own numbers tell you exactly where the leak is. The trick is reading the ratios, not the totals. Here is what a healthy funnel from a well-targeted list of 1,000 looks like, so you can spot where yours diverges.
A healthy LinkedIn funnel from a well-targeted list of 1,000.
Now read your own against it. The diagnosis is usually one of three things:
- Low acceptance (under 25%). The problem is upstream: your targeting, your profile, or your connection note. The message after connecting cannot fix a list that never accepts.
- Good acceptance, low replies. People accept but do not engage. That is a message problem. Your opener is generic or your timing is off.
- Good replies, few positive ones. People answer but say no. That is a targeting precision problem. You are reaching the wrong people, or the right people at the wrong moment.
One caution that does not show up in the funnel: if your numbers suddenly crater across the board, check account health before you rewrite anything. LinkedIn throttles profiles that look automated, and a restricted account quietly tanks every metric at once. We cover the warning signs in our piece on restricted LinkedIn accounts. Running outreach on a dedicated account, paced like a human, is how you keep that variable out of the equation entirely.
Key takeaways
- 2026 averages: ~28 to 30% acceptance, 10.4% reply rate after connecting. Good is 35%+ and 25%+.
- Track positive replies and meetings booked, not raw reply rate. Strong is 5 to 8% positive of sends.
- Connection-note replies fell 37% in a year because generic templates flooded the channel.
- Benchmark against your own industry: recruiting runs 18 to 25% replies, software 8 to 9%.
- The levers that beat the benchmark: a real message, signal timing, follow-ups, brevity. Not volume.
- Read the ratios to find your leak: low acceptance is targeting, low replies is the message, low positives is precision.