What LinkedIn InMail Costs Per Reply

A Sales Navigator Core seat is US$119.99 a month and comes with 50 InMail credits, which makes the cost per send look like US$2.40. That number is wrong in both directions, and the ways it is wrong are the whole story of whether this channel is worth running.

It is too high because LinkedIn refunds the credit whenever someone responds, so a seat that gets replies buys more than 50 sends. It is too low because a send is not the thing you are buying. Replies are. And the gap between cost per send and cost per reply is set entirely by targeting, which is the variable most teams treat as fixed.

What you are actually paying

Sales Navigator pricing, verified on LinkedIn's own compare-plans page:

Prices exclude tax and are per licence. The credit column is not a typo: LinkedIn states that all Sales Navigator subscription types provide the same number of credits, so the extra US$40 a month for Advanced buys research and collaboration features rather than messaging capacity. Our Sales Navigator pricing guide covers what that money does buy.

There is no way to purchase additional credits on these plans. LinkedIn is explicit about it, which means the seat cost is the entire InMail budget and volume is capped by subscription count rather than by spend.

The refund rule changes the denominator

LinkedIn credits back every InMail that is accepted, declined, or responded to directly within 90 days of being sent, immediately on the recipient's response. Auto replies count. Quick replies count. A "Not interested" is recorded as declined and still refunds the credit.

So the credits you consume in a month are the messages that were ignored, and the messages that worked were free.

1. Step 1Fixed input

   One seat, one monthly fee, 50 credits allocated on the first of the month.

2. Step 2Sends are variable

   Every response refunds a credit, so a higher response rate buys more sends from the same allocation.

3. Step 3Replies are what you wanted

   The seat cost divided by replies is the number that maps to pipeline.

4. Step 4One lever moves both

   Better targeting raises replies and lowers consumption at the same time.

The arithmetic, worked

The figures below are illustrative arithmetic on a US$119.99 Core seat, not measured results. Substitute your own response rate and the shape holds.

Take a seat that sends its full allocation each month and treat everything as a single monthly cycle for simplicity.

At a 10% response rate: 50 sends produce 5 responses. Five credits return, so you can send 5 more, which at the same rate produces roughly half a response. Call it 55 sends and 5 replies. The seat cost divided by replies is about US$24 per reply.

At a 25% response rate: 50 sends produce roughly 13 responses. Those refunds fund another 13 sends, which produce another 3. Call it 66 sends and 16 replies, or roughly US$7.50 per reply.

At a 4% response rate: 50 sends produce 2 responses. Effectively 52 sends and 2 replies, at roughly US$60 per reply.

25% response rate~US$7.50 per reply

Roughly 66 sends, 16 replies, credits largely self-refunding

10% response rate~US$24 per reply

Roughly 55 sends, 5 replies

4% response rate~US$60 per reply

Roughly 52 sends, 2 replies, allowance emptied

A sixfold difference in cost per reply, on identical spend, driven by nothing except who you chose to message. That is unusual. In most paid channels a bad campaign costs the same as a good one and returns less. Here a bad campaign costs more per outcome and consumes more of the fixed resource while doing it.

The costs that are not on the invoice

Three of them, and they are frequently larger than the subscription.

The time to target properly. The arithmetic above says targeting quality is the only variable that matters, and targeting quality is human hours. Someone has to read profiles, verify roles are current, and find the specific reason to write today. On a 50-message allowance that is real work per message, and it is the work people skip when volume feels like the goal.

The one-shot constraint. A LinkedIn thread is permanent and visible, so a second message lands directly beneath the one they ignored, reading as a bump whatever the campaign structure calls it. We do not run no-reply retargets on LinkedIn for that reason. Practically, it means a wasted first touch is a permanently closed door on that channel for that person, which is a cost email simply does not carry.

Account risk. Activity limits on invitations, searches, profile views and messaging apply to the account regardless of subscription, and exceeding them puts the account itself at risk rather than just queuing the work. The LinkedIn automation tools guide covers where those sit, which matters most when a tool is sending on your behalf.

When the honest answer is not to buy the seat

Three situations where the arithmetic says no, and it is worth recognising them before the renewal rather than after.

Your buyers are not active on LinkedIn. Some markets genuinely are not there. Trades, field operations, much of manufacturing and a good deal of local services have thin LinkedIn presence among the actual decision makers, and no amount of credit efficiency fixes an audience that does not read the platform.

Nobody owns the research time. The seat only works if someone spends real minutes per message. A seat bought by a team with no capacity to use it carefully becomes 50 generic messages a month, which is the configuration with the worst cost per reply available.

You need the volume this quarter. Fifty messages a month does not build a pipeline on a deadline. If the requirement is a step change in meetings within a quarter, the seat is not the instrument, and buying it delays the conversation about what is.

None of those make LinkedIn a bad channel. They make it the wrong first purchase, and the sequence that usually works is a functioning email programme first, with LinkedIn added for the accounts email cannot reach.

How it compares to email

Cold email costs a fraction of a cent per send in infrastructure and does not cap at 50 a month. On raw cost per touch it is not a contest.

Where InMail earns its price is the population that email cannot reach effectively: buyers whose inboxes are saturated, whose addresses are hard to verify, or who are demonstrably more responsive on LinkedIn. For those accounts the relevant comparison is not against the cost of an email but against the cost of not reaching them.

The practical model is email for coverage and InMail for the accounts where coverage is not working, with LinkedIn volume kept deliberately small. If you are benchmarking the email side, our cold email reply rate benchmarks give the comparison figures.

Running the calculation on your own numbers

Four inputs, and you can do it on paper in five minutes.

Start with the annualised seat cost. Core at US$1,079.88 a year or Advanced at US$1,799.88, per licence, excluding tax. Monthly billing is more expensive on both, so use whichever you actually pay.

Add the loaded cost of the hours. Whoever researches and writes the messages costs something per hour, and at 50 messages a month even ten minutes each is over eight hours. On many teams this exceeds the subscription, which is the finding that usually changes the decision.

Estimate sends, including refunds. Take your allocation, multiply your response rate by it to get refunded credits, and add those back. A 20% response rate on 50 credits gives roughly 60 sends in a month before second-order effects.

Divide by replies, then by meetings. Total monthly cost divided by replies gives cost per reply. Divided by meetings booked, it gives the number that is actually comparable to your other channels.

The result is usually uncomfortable at low response rates, which is the point of doing it. A channel that costs US$60 a reply and converts a fifth of replies into meetings is a US$300 meeting before anyone has spoken to a prospect.

What to actually measure

Cost per credit is not a metric. Cost per reply is a decent one. Cost per qualified meeting is the one that decides whether the seat stays.

Track sends, responses inside 90 days, and meetings booked from those responses, per seat and per month. If cost per meeting on LinkedIn is materially worse than on email, the answer is almost never more seats. It is fewer, better-chosen messages, because that is the only input the arithmetic responds to.

For the underlying mechanics, how InMail credits work covers the refund and expiry rules in detail, and what InMail is and how it differs from a connection request covers when a credit is the right thing to spend at all.

We run LinkedIn and email together for clients on a pay-per-qualified-meeting basis, which moves this arithmetic off your invoice entirely: we carry the seat cost and the targeting hours, and you pay for meetings. You can see what a campaign would look like for your market.

Sales Navigator pricing is per LinkedIn's compare-plans page as of August 2026 and excludes VAT, GST and promotional discounts. Credit allowances and refund rules are per the LinkedIn Help Center. Cost-per-reply figures are illustrative arithmetic, not measured results. Verify current pricing with LinkedIn before budgeting.

Sources: Sales Navigator compare plans, Understand InMail credits in Sales Navigator, InMail message credits and renewal process