A founder we work with had a post clear 90,000 impressions in April. Best post of his life. Then Q2 planning ate his calendar, he went dark for five weeks, and in June he came back with a post that was — honestly — better than the April one. Sharper hook, better receipt, tighter close.
It died at 340 impressions.
He assumed the algorithm was punishing him. It wasn't, exactly. His June post got graded the same way every post gets graded. What changed was where the grading started. The April post launched off a warm account with an eight-week posting streak behind it. The June post launched off a cold one.
That's the piece most founders never see: LinkedIn runs two ledgers. One scores the post. The other scores the account — and the account ledger decides the size of the audience your post gets tested on before it earns anything at all.
The Two Ledgers: Post Score and Account Baseline
We've written before about how the test-pool system grades individual posts — small sample, dwell and comments, promote or cap. That's ledger one, and it resets with every post.
Ledger two doesn't reset. It's an account-level quality expectation — call it an authority score — that LinkedIn's 360Brew system maintains on you as an author. Current platform analyses describe it directly: creators who post consistently, in a consistent topical lane, build a score that raises the baseline reach of every subsequent post. The observable pattern across accounts is that 3+ posts per week sustained for 8+ consecutive weeks is roughly where the floor starts moving.
Think of it as the difference between a single listing's conversion rate and your account health score. One is per-asset. The other follows you around and changes how the platform treats everything you ship.
The practical effect: two identical posts from two accounts get different opening test pools. The account with the higher baseline starts its exam in front of 3-5x more people. Same post, different starting line.
Why a Viral Post Doesn't Raise Your Floor
Here's the part that stings: the 90K-impression post did almost nothing for his baseline.
A spike is a statement about one post. The baseline is a prediction about your next post — and the algorithm builds that prediction from reliability, not peaks. One outlier tells 360Brew very little about what you'll publish Tuesday. Twenty-four consecutive weeks of solid, on-lane posts tells it a lot.
This is why "I went viral once" founders and "I post three times a week" founders have completely different distribution curves twelve months in. The viral founder is re-auditioning from a cold start every time. The consistent founder walks into a bigger room before saying a word.
It also explains a pattern we see constantly in ecommerce founder accounts: the follower count and the reach stop correlating. A 40K-follower account that posts in bursts gets out-reached by an 8K-follower account with an unbroken cadence. Followers are ledger zero — the algorithm barely checks it anymore. The interest graph and the author baseline do the routing.
The Dark Month Costs More Than the Bad Post
Founders optimize the wrong risk. They'll sit on a draft for a week polishing it because a mediocre post feels dangerous. Meanwhile they'll skip three weeks during inventory season without a second thought.
The math runs the other way. A mediocre post caps early in the test pool — a contained, per-post loss that the account ledger barely notices. A dark month is an account-level withdrawal. The streak breaks, the baseline decays, and the next six posts launch into shrunken test pools regardless of how good they are.
That's the founder from the intro. His June post didn't fail on quality. It paid the tax on five silent weeks — and the two or three posts after it paid the same tax while the baseline rebuilt.
For ecommerce founders specifically this is brutal, because the busy seasons — Q4 prep, Prime Day, freight crises — are exactly when posting stops. The account goes cold at the moments the founder has the best material.
Protecting the Baseline When You're Slammed
You don't need heroic output to keep the account ledger warm. You need to not go to zero.
Set a minimum viable cadence, not a maximum. Three posts a week sustains the streak. When the quarter goes sideways, drop to the floor — don't drop to nothing. Two lighter posts beat zero perfect ones by a wide margin, because the ledger is scoring presence and reliability, not per-post brilliance.
Keep a break-glass reserve. Four or five evergreen posts — a framework, a question you get on every sales call, an old post reframed — written in a calm week and held for chaos weeks. Their job isn't to perform. Their job is to keep the streak alive so your good posts land on a warm account.
Stay in your lane while you do it. The baseline is topical, not generic. A filler post about your niche maintains it. A random off-lane post about productivity hacks doesn't — and can actively muddy the interest-graph signal the baseline is built on.
Decouple writing from publishing. This is the structural fix. Founders who write-then-post are one bad week from a broken streak. Founders with a two-week scheduled buffer are not. Cadence stops being a willpower question and becomes an inventory question — and ecommerce founders are good at inventory.
What This Changes About How You Read Your Analytics
Stop judging posts in isolation. Before you declare a post a failure, check what the account looked like when it launched.
A post that flops during an active streak is a content problem — wrong hook, wrong topic, wrong day. Fix the post.
A post that flops after a gap is a baseline problem. The content might be fine. Reposting the same material after four weeks of restored cadence is a legitimate test — we've watched near-identical posts go from ~400 impressions to 4x-6x that on nothing but a warmer account.
And when you return from a dark stretch, expect the first two or three posts to underperform no matter what. That's not the algorithm rejecting you. That's the rebuild. Founders who don't know this quit again right before the baseline recovers — which restarts the whole cycle.
FAQ
Does posting daily build the baseline faster than 3x a week? Marginally, and it usually isn't worth it. The steep gains are between zero and consistent — not between consistent and heavy. Daily posting from a founder who can't sustain it sets up a visible frequency crash, which is worse than a steady three.
How long does it take to rebuild after going dark? Plan on 4-8 weeks of restored cadence before the floor recovers, consistent with the 8-week streak pattern. It rebuilds faster than it built the first time — the topical history doesn't vanish — but it is not instant, and the first posts back will look discouraging.
I went viral last month. Why is my reach normal again? Because the spike was a post-level event, not an account-level one. Nothing is broken. The only way to convert a viral moment into a durable floor is to keep publishing on-lane through the weeks after it — that's when the reliability signal actually compounds.
Does the baseline transfer if I change topics? Mostly no. The score is built on topical consistency, so a hard pivot puts you closer to a cold start than founders expect. If you're repositioning, transition gradually and keep cadence high while the interest graph re-learns you.
This is the actual case for ghostwriting, by the way — not "better writing." Most founders can write one great post. Almost none can ship three decent ones a week through Q4, a 3PL fire, and a product launch. The baseline doesn't care which weeks were hard. If your account has been running hot-and-cold and your best posts keep launching into empty rooms, that's the pattern we fix first. Get in touch and we'll look at your last 90 days together.