The measurement guide · four numbers, honestly

LinkedIn analytics for founders: the four numbers that matter and the six that don't

Ezekiah Livingston LinkedIn strategist and ghostwriter. I refuse to promise engagement on principle, which makes me picky about which numbers deserve your attention. Find me on LinkedIn.

Published August 28, 2026

TL;DR

Four numbers matter: the 90-day impressions trend, saves plus substantive comments, profile views after posts, and inbound conversations from people who could actually buy.

Six don't: likes, follower count, one viral post, connection count, SSI, and engagement rate on tiny denominators.

Measure in 90-day windows. Weekly analytics is anxiety with a chart.

The question behind every metric: did the right stranger take a next step?

LinkedIn hands founders a dashboard designed for creators, then founders wonder why the numbers never connect to revenue. Most of that dashboard is noise for your purposes. Four numbers deserve ten minutes a month; the rest deserve to be ignored in good conscience.

The four that matter

1. The 90-day impressions trend. Not this week against last week, which is noise, but the quarter's direction. It's your baseline health check, and when it turns down and stays down, the reach diagnostic tells you which of the four causes you have.

2. Saves and substantive comments. A save is a bookmark: someone decided your post has value past this scroll. A comment longer than a sentence is a conversation. Both are depth signals the ranking model appears to reward, and neither can be faked cheaply.

3. Profile views following posts. This is your content-to-curiosity conversion. Posts that spike profile views are pulling the right kind of attention, and then your profile has to convert it, which is exactly what the free 8-Second Test checks.

4. Inbound from people who could buy. DMs, replies, booked calls from your actual market. The only number revenue recognizes. Five of these a quarter beats five thousand impressions a week, and it isn't close.

The six that don't

Likes are the cheapest signal on the platform and price in accordingly. Follower count without buyer density is an audience of peers applauding each other. One viral post is a lottery ticket, not a strategy, and its impressions mostly land outside your market. Connection count stopped meaning anything at the mass-connect era. SSI is a Sales Navigator artifact that measures tool usage, not influence. And engagement rate on a 40-follower denominator is arithmetic cosplay. None of these numbers can survive the question "so what did it cause?"

The ten-minute monthly dashboard

Once a month, four rows in a spreadsheet:

Metric	Where	What a problem looks like
Impressions, trailing 90 days	Creator analytics	Down two months running: run the reach diagnostic
Saves + real comments per post	Count your last 8 posts	Falling while impressions hold: depth problem, fix substance
Profile views	Profile analytics	Flat while posting: hooks or topic problem
ICP conversations started	Your inbox, honestly counted	Zero for a quarter: profile isn't converting, test it

Ten minutes, once a month. More often than that and you're day-trading your own attention.

Why I refuse to promise engagement

Every number above measures distribution and response, and nobody controls those, only the craft that earns them. That's why the free Content Audit scores craft and explicitly refuses to predict engagement: well-written and well-distributed are different claims, and anyone promising the second one is guessing with your money. The AI vs human data shows what happens when accounts optimize for output volume instead.

The gray-market test: if a number can be bought for $50 from an engagement panel, it is not a business metric. Likes, followers, and comment pods all fail that test. Saves, profile views from strangers, and booked calls pass it.

Free · two minutes

Your profile is where the numbers convert

Impressions become revenue through one page: your profile. The 8-Second Test shows the verdict a buyer forms before deciding to message you.

Test my profile Audit my posts

Questions people ask

What is a good engagement rate on LinkedIn?

Mostly the wrong question. Engagement rate on small follower counts is statistically meaningless, and likes, its main ingredient, are the platform's cheapest signal. Track depth per post instead: saves, comments longer than a sentence, and profile views that follow publishing.

How often should I check LinkedIn analytics?

Monthly, for about ten minutes, across four numbers: the 90-day impressions trend, saves plus substantive comments, profile views, and inbound conversations from your market. Weekly checking measures your anxiety, not your content.

Do LinkedIn impressions matter?

As a 90-day trend, yes: it's the baseline health metric, and a sustained decline means it's time to diagnose. As a daily or weekly number, no: normal volatility swamps the signal, and single-post spikes mostly reach people outside your market.

What is a good SSI score?

The Social Selling Index is a Sales Navigator artifact that largely measures how much you use LinkedIn's sales tooling. It has no meaningful relationship to whether buyers trust your profile or your content. Ignore it without guilt.

How do I know if my LinkedIn content is actually working?

Two numbers, ninety days: profile views following your posts (content is pulling the right curiosity) and conversations started with people who could buy, hire, or refer (attention is converting). If the first rises and the second doesn't, the leak is your profile, and it's testable free.

Keep going: the reach-drop diagnostic, how to test your profile, or the hook-writing guide if depth per post is the weak row.