Guides
How to check your LinkedIn B2B reach algorithm without a dashboard
LinkedIn B2B reach algorithm checks need one number you can defend. Here is the metric, its limit, and the exact threshold for changing course today.
What to take away
- Track impressions per follower over a rolling 28 days, not raw impressions. The ratio is comparable across account sizes.
- Read the number against your own trailing baseline, not against another company's post.
- Act when the ratio drops 30 percent or more for two consecutive weeks. Below that, the move is noise.
- The metric cannot separate a ranking change from weak creative or a holiday week.
- LinkedIn does not publish per-post ranking weights, so the ratio is a proxy, not a diagnosis.
LinkedIn's feed does not hand you a dashboard for organic B2B reach. The platform shows impressions, members reached, and engagement, but nothing that labels a ranking shift. You build the instrument yourself, and you build it from numbers the platform already gives you.
What to measure
Define the metric first. Impressions per follower is total organic impressions on posts published in a period divided by follower count at the start of that period. A 12,000 follower page that draws 36,000 impressions over 28 days sits at 3.0. That single figure travels across quarters and across accounts of different sizes.
Pull the inputs from the page analytics export, not from the feed. Sort posts by publish date, sum impressions for the window, and divide by the follower count on day one. Rebuild the same figure for the previous four windows so you have a baseline.
| Window | Impressions | Followers at start | Impressions per follower |
|---|---|---|---|
| Days 1-28 | 31,000 | 12,000 | 2.6 |
| Days 29-56 | 36,000 | 11,800 | 3.1 |
| Days 57-84 | 34,500 | 11,900 | 2.9 |
| Days 85-112 | 22,000 | 12,000 | 1.8 |
The last row is the one that matters. A drop from 2.9 to 1.8 is a 38 percent slide, well past the threshold below.
How to read it
Compare the current window to the trailing four-window average, not to a single prior month. A one-month comparison will fire on a quiet news week. The four-window average absorbs holidays, conference seasons, and the odd post that travels.
- Export 28 days of post impressions and the follower count on day one
- Compute impressions per follower for the current window
- Compute the same figure for the prior four windows
- Compare current against the four-window mean
- Flag the window only if the gap clears 30 percent
Set the threshold at 30 percent for two consecutive weeks. Below that, the movement sits inside normal variance for most B2B pages, and a change in posting cadence will cost you more than it recovers. Above it, treat the drop as real and audit the obvious causes before blaming ranking.
What it cannot tell you
The metric cannot separate a ranking change from a content change. If you shifted from document posts to text-only posts in the same window, the ratio moves for reasons that have nothing to do with distribution. The number is blind to format, posting hour, and audience saturation.
It also says nothing about who saw the post. Impressions count views, not qualified buyers. A reach drop among followers who never enter your pipeline is not a business problem, and a reach gain among the same group is not a win. For a wider view of how reach behaves across platforms, see viral reach properties, which lays out the signals that decide whether strangers see a post at all.
Attribution and its limits
Attribution on LinkedIn runs through a form fill, a meeting booked, or a tracked link. The reach metric sits upstream of all of them, and the chain breaks in the middle. A prospect can see three posts, ignore the link, and arrive through a colleague's forward. Nothing in the analytics records that path.
The platform itself is a moving target. Wikipedia's overview of LinkedIn notes that the feed's ranking has shifted repeatedly since the platform introduced algorithmic sorting, which means your baseline ages faster than a quarterly report suggests. When the ratio moves, you are measuring a platform you cannot inspect and a market you cannot isolate.
US state privacy laws add a second layer. If your targeting depends on inferred data, the rules differ by state, and California, Colorado, and Virginia each impose their own notice and opt-out duties. The Federal Trade Commission's guidance on social media disclosures applies to how you describe sponsored reach, which affects what you can claim about a campaign's results.
When to stop measuring and decide
Run the check weekly for eight weeks. If the ratio holds above the threshold, keep the cadence and stop auditing. If it clears the 30 percent drop for two straight weeks, change one variable: posting format or posting time, not both.
A single change gives you a clean read on the next window. Two changes at once leave you unable to attribute the recovery. Document what you changed and the date, so the next window has a reference point.
For a method that catches quiet ranking shifts before they show in reach, see algorithm change tracking tools compared. It covers what actually detects an update rather than what claims to.
Common questions
Why use impressions per follower instead of total impressions? Total impressions rise with follower count, so a growing page looks healthy even when distribution weakens. The ratio controls for size and makes windows comparable.
What if my page is small? Below roughly 2,000 followers, weekly swings are large and the 30 percent threshold fires often. Use a 56-day window instead of 28 days and judge on the longer average.
Does a reach drop mean the algorithm penalized me? Not necessarily. Format, timing, and audience fatigue produce the same pattern. The metric tells you something changed, not what changed.
How often should I rebuild the baseline? Every quarter. Older windows drift as the platform's ranking evolves and as your follower mix shifts, so a stale baseline will misread a normal month as a decline.
