The Two LinkedIn Numbers Worth Watching.

Open the analytics on any LinkedIn post and you get a row of numbers: impressions, members reached, reactions, comments, reposts.

Two of them are worth more than the rest put together, and most teams never scroll far enough to read them.

Saves and Sends. LinkedIn’s own documentation defines them plainly. Saves is the number of times members saved your post. Sends on LinkedIn is the number of times members sent your post to someone else. Both sit in the engagement breakdown of post analytics, next to the reactions and comments that get screenshotted for the monthly update.

They are worth more because they cost the reader something. A like is a reflex. A save is a decision to keep something. A send is a decision to attach your post to a private message with a colleague’s name on it.

Sort a month of posts by those two columns and you stop looking at a scoreboard and start looking at a content plan.

What Saves and Sends actually count

A save is a bookmark. Someone taps the icon and your post goes into their saved items, where they can find it again without searching for you.

A send is a private share. Someone picks your post, picks a person or a small group, and passes it into a LinkedIn message thread.

LinkedIn describes both as a number of times, not a number of people, so read them as volume of action rather than headcount. Neither one notifies you individually. You only ever see the total.

That privacy is the whole point. Sends in particular describe a distribution path that never touches the public feed, which is why the number was invisible in social reporting for so long.

Why the public numbers stopped telling you much

Likes on LinkedIn are close to free. They are one tap, they are often reciprocal, and a decent share of them come from people who engaged with your last post rather than people who read this one.

Comments are worse, because they can be manufactured. Engagement pods, tag chains and a run of “great post” replies all produce a number that looks like interest and contains none. The broader case for judging social by metrics that actually matter for brand growth applies here with force.

On LinkedIn the problem is sharper than usual. The platform rewards commenting, so the metric that is easiest to inflate is also the one the feed pushes you to chase.

Saves and Sends have no such economy behind them. Nobody saves a post to be polite.

A save is a bet the post will be needed again

Saving is a filing decision. The person is not applauding you, they are telling themselves they will want this back.

That makes Saves a reliable measure of utility. Posts that get saved tend to be the ones carrying something reusable: a framework, a checklist, a set of numbers, a process someone intends to copy on Monday.

It also means a low save count on a well-liked post is information, not failure. The post entertained people. It did not give them anything to keep.

If your commercial goal is to be the account a buyer returns to when a problem lands on their desk, Saves is the closest thing LinkedIn gives you to a measure of that.

A send is someone putting their name to it

Forwarding a post in a private message is a small act of endorsement. The sender has to choose a recipient, and they carry the risk of wasting that person’s time.

Nobody sends a weak post to their boss.

Our reading is that a send is the strongest endorsement the platform offers, because it is the only engagement that costs the person social capital inside their own organisation. A like is performed for an audience. A send is performed for one colleague who will judge it.

It is also the engagement most likely to sit upstream of a conversation. When a post reaches the person who owns the budget, it tends to arrive by message rather than by feed.

You cannot see who sent it or to whom, so treat the number as a direction, not a lead list.

Where the two numbers live

Saves and Sends appear in the engagement section of analytics on an individual post, alongside reactions, comments and reposts.

They also appear in combined post analytics, which LinkedIn documents as available to all members from the Analytics and tools section. That view covers a date range from seven to three hundred and sixty five days and exports to a spreadsheet.

The export is what turns this from a curiosity into a workflow. A column you can sort beats a number you squint at.

Two honest caveats. Interfaces move, so check the current layout in your own account before you write a process document around it. And if you run a company Page rather than a personal profile, confirm the metrics are present in your admin analytics view rather than assuming parity, because LinkedIn’s Page-side documentation describes reactions, comments and reposts without naming Saves and Sends.

The reach line that sits next to them

LinkedIn also splits your impressions by whether the viewer follows you or is connected to you, reported as in network and out of network. Members reached is held for a shorter window than the engagement data, so pull it while it is there.

Pair that split with Saves and Sends and you can answer two separate questions at once: did this post reach anyone new, and did anyone value it.

• High out of network reach, low saves and sends. The packaging travelled and the substance did not. Fix the body, keep the hook.
• High saves and sends, low out of network reach. Genuinely useful to people who already know you. Rewrite the opening for a colder reader and post it again in a different form.
• Both high. This is the topic. Do more of it and go deeper, not wider.
• Both low. Either the topic is wrong for this audience or the first two lines lost them. Check the hook before you retire the subject.

Most teams read only the first number in that pair, which is how a brand ends up celebrating reach it did not want.

The monthly review that turns this into a plan

This takes about twenty minutes and replaces most of what a LinkedIn report currently contains. It also slots neatly into whatever review cadence you already run, or into the wider diagnostic work of a social audit and strategy engagement.

1. Export the last thirty days of posts from combined post analytics.
2. Sort by Sends, highest first. Write down the top five topics, not the top five posts.
3. Sort by Saves, highest first, and write down that top five as well.
4. Mark the overlap. Topics on both lists are the ones your audience trusts you on.
5. Check the bottom of both sorts. Those are the formats to stop producing, however well they perform on likes.
6. Build next month’s plan from the overlap, then take those two lists into the leadership update instead of an impressions chart.

Do it monthly, not weekly. Thirty days is about the smallest window that yields enough posts to show a pattern rather than a fluke.

Read topics, not posts

The single most common error is treating one high-send post as a verdict. Sends on any individual post are a small number and small numbers are noisy.

The pattern across a month is the useful object. Three posts about pricing that all got sent tell you something. One post that got sent nine times tells you almost nothing until it repeats.

Do not go looking for a benchmark save rate either. There is no public figure worth trusting, and comparing your account to somebody else’s average is a good way to make a bad decision.

Your own baseline over ninety days is the only comparison that means anything. Track the trend, not the absolute.

The mistakes that ruin the read

• Asking for it. “Save this for later” inflates the number and destroys the thing that made it useful, which is that nobody had to be prompted.
• Reporting the count without the topic. Twelve sends is not a finding. Twelve sends on hiring posts is.
• Treating a send as a lead. You cannot see the recipient, and pretending otherwise puts a fiction in your reporting.
• Letting the metric flatten your content into reference material only. Saves reward utility, which quietly punishes point of view, and a feed with no opinion in it stops being read at all.
• Changing your mind every month. If the same topics keep surfacing, the answer is more depth on them, not a new experiment.

What good looks like

A team with this working can name, without opening a dashboard, the three topics their audience keeps saving and the two that colleagues forward to each other.

Their monthly LinkedIn review is one page: the two sorted lists, the in network and out of network split, and one decision about what changes next month.

Their content plan is visibly downstream of that. New posts extend the topics that earned sends rather than restarting from a brainstorm.

And their reporting to leadership leads with what the audience valued, then supports it with reach, rather than the other way round. The conversation shifts from how many people saw it to what the people who saw it did next.

How NBK thinks about LinkedIn measurement

Reporting is part of the operating system, not a task you do at the end of the month. If the numbers a team reads do not change what they make, the reporting is decoration.

Saves and Sends are useful precisely because they are hard to argue with in a meeting. They point at topics, and topics are actionable in a way an engagement rate never is.

That is the same logic behind treating founder and executive posting as a system rather than a hobby, which is why personal brands need a LinkedIn content system with a review step built into it. A metric only earns its place if something happens as a result of reading it.

None of this needs a new tool or a bigger team. It needs somebody to open the export, sort two columns and act on what they see.

Next step

If your LinkedIn reporting is busy but never changes the plan, start by sorting your last thirty days by Sends and seeing whether the answer surprises you.

If it does, and the content calendar still looks nothing like that list, NBK can help find the constraint in the system behind it.