Quick answer: The four LinkedIn outreach metrics that matter most are acceptance rate (accepted divided by sent), reply rate (replies divided by messages delivered), positive reply rate (interested replies divided by messages delivered), and meetings booked. Track them in order, because each depends on the one before. Build your own baseline rather than copying someone else's benchmark.
Why these four
Outreach is a funnel. People move through stages, and each stage has its own number:
- You send a connection request.
- Some accept.
- You message those who accepted.
- Some reply.
- Some of those replies are interested.
- Some interested people book a meeting.
Metrics on vanity activity ("messages sent") tell you what you did. Metrics on conversion tell you what worked. Focus on the second.
The four metrics
| Metric | Formula | What it tells you |
|---|---|---|
| Acceptance rate | Accepted requests / requests sent | Whether your targeting and profile are convincing |
| Reply rate | People who replied / people who received a message | Whether your messaging is getting attention |
| Positive reply rate | Interested replies / people who received a message | Whether the right people want what you offer |
| Meetings booked | Count of meetings (or meetings / people messaged) | Whether the funnel produces outcomes |
Use these illustrative numbers to see how it works. They are made up for the example. Suppose you send 200 connection requests, 60 accept, you message those 60, 12 reply, 5 of those are interested, and 3 book a meeting.
- Acceptance rate: 60 / 200 = 30%
- Reply rate: 12 / 60 = 20%
- Positive reply rate: 5 / 60, about 8%
- Meetings: 3 (that is 3 / 60 = 5% of people messaged, or 1.5% of requests sent)
Always write down the denominator. Replies divided by requests sent and replies divided by messages delivered are different numbers, and mixing them causes arguments that are really about definitions.
Define each term before you count
Agree on these definitions with your team and keep them fixed:
- Sent: a request that actually went out, not one queued or blocked.
- Accepted: the person accepted the request, not just viewed the profile.
- Reply: any response from the person, not an automatic message.
- Positive reply: a reply showing interest or willingness to talk. Decide how you will classify "not now" and "send info."
- Meeting booked: a confirmed meeting. Consider tracking "held" separately, since no-shows exist.
Diagnosing a low number
Each metric points to a different part of the funnel, so a problem in one place calls for a specific fix.
Low acceptance rate
Likely causes:
- Targeting is too broad or off-profile
- Profile is thin, generic, or does not match your claim
- The connection note reads as a pitch
- Volume is too high for the account
Try: tighten the list, improve the profile, test blank versus short specific notes (see connection request message examples), and lower the pace. A falling acceptance rate is also a safety signal; see how many connection requests per day.
Low reply rate (with healthy acceptance)
Likely causes:
- The first message pitches too early
- It is long, generic, or full of jargon
- The ask is too big
- Timing or follow-up is weak
Try: shorten, make it specific, ask one small question, and add follow-ups. See how to write cold LinkedIn messages that get replies and LinkedIn follow-up templates.
Replies but few positive replies
Likely causes:
- The list reaches people who are not a fit
- Your offer is unclear or not relevant to their role
- Timing is wrong for the market
Try: revisit who you are targeting, refine the offer, and read the replies for patterns in objections.
Positive replies but few meetings
Likely causes:
- Slow response to interest
- Friction in booking
- The next step is too big
Try: respond quickly, offer easy time options, and simplify the ask. Speed is often the biggest lever here.
How to build your own baseline
Published benchmarks vary enormously by industry, role, list quality, and message, so use them only as a loose reference, if at all. A baseline from your own data is far more useful.
- Pick one audience and one sequence. Don't mix very different groups.
- Collect enough volume. A few sends can mislead; keep going until the numbers are stable across several batches.
- Record each metric weekly with its denominator, not just a total.
- Set a range, not a point. Your baseline might be "acceptance between X and Y." Weekly numbers bounce.
- Change one thing at a time. If you alter the list, the note, and the follow-up together, you will not know what moved the number.
- Compare like with like. Same audience, same period, same account type.
- Review every month. Seasons, news, and platform changes shift results.
Watch for these traps
- Chasing volume. More sends with falling acceptance is moving backward and can raise account risk.
- Counting activity as results. Messages sent is not an outcome.
- Small-sample conclusions. One great week is an anecdote.
- Ignoring lag. Replies and meetings arrive days after sending, so measure a cohort after enough time has passed.
- Blending channels. If you also use email, report each channel separately, then together. See LinkedIn vs email outreach.
- Tracking only success. Also record warnings, refusals, and restrictions on the account; see the weekly invite limit explained.
Where FollowArc fits
FollowArc runs outreach across LinkedIn and email and ranks every reply by intent, which makes the reply and positive-reply numbers easier to judge without sorting by hand. We do not publish benchmarks, because results depend too heavily on the audience; the useful numbers are the ones from your own campaigns. To build a baseline on your own list, try FollowArc free for 7 days or read about the AI SDR use case.