How LinkedIn Actually Works: Algorithm, Content, and Traction

LinkedIn is not a social network where you post and hope. It is an algorithm-driven distribution system with specific rules, measurable signals, and predictable patterns. Understanding those rules is the difference between shouting into the void and building an audience that brings you opportunities.

Who this is for

You have a LinkedIn profile. Maybe you post occasionally, maybe you just scroll. You have noticed that some people seem to get thousands of views while your posts get 47. You suspect there is a system behind it but you do not know how it works.

This path is for you if:

You want to build professional visibility and attract opportunities (clients, partnerships, hires, collaborators)
You are building a brand, a product, or a practice and want LinkedIn to be part of your strategy
You have tried posting and found the results inconsistent
You want to understand why things work, not just copy tactics

What this path is NOT

This is not a list of growth hacks. Growth hacks expire when the algorithm changes. This is a structural understanding of how LinkedIn distributes content, what signals it rewards, and how to build a sustainable presence. The algorithm changes every year. The principles behind it change much less.

Part 1 — The machine behind the feed

LinkedIn is not showing you everything your connections post. It is showing you what its algorithm predicts you will engage with. Understanding that prediction system is the foundation of everything else in this path.

How the feed is assembled

Richard van der Blom’s Algorithm Insights Report 2025 — the most comprehensive independent study, analysing 1.8 million posts from 58,000 profiles — reveals how the feed is composed:¹

graph LR
    subgraph Feed Composition
        TC[Top Creator<br/>Content<br/>31%]
        OC[Other Creator<br/>Content<br/>28%]
        PC[Promoted Company<br/>Content<br/>28%]
        AD[LinkedIn Ads<br/>11%]
        OG[Organic Company<br/>Pages<br/>2%]
    end

    style TC fill:#4a9ede,color:#fff
    style PC fill:#f0ad4e,color:#fff
    style OG fill:#d9534f,color:#fff

The critical number: organic company page content accounts for just 2% of the feed. Personal profiles dominate. This is not an accident — it is a deliberate design choice by LinkedIn to prioritise people over brands.¹

The four-stage pipeline

Every post you publish passes through a multi-stage evaluation before it reaches its full audience:²

graph TD
    P[Post Published] --> Q[Quality Filter<br/>First 60 minutes]
    Q --> G[Golden Window<br/>First 90 minutes]
    G --> R[Extended Review<br/>8 hours]
    R --> F[Final Distribution<br/>24 hours]
    Q -.->|Low quality| D[Suppressed]

    style G fill:#4a9ede,color:#fff
    style Q fill:#f0ad4e,color:#fff

Stage	Timeframe	What happens
Quality filter	First 60 min	Post clears a quality gate and is pushed to a small initial audience. The algorithm watches early engagement.
Golden window	First 90 min	Most of your reach is determined here. Early engagement signals tell the algorithm whether to push wider.
Extended review	8 hours	Continued monitoring of conversation depth, engagement quality, and consumption metrics.
Final distribution	24 hours	Posts that sustain engagement get extended distribution beyond your network.

This is why the first 90 minutes after posting matter more than everything else combined. A post that gets meaningful engagement in the golden window will be shown to a wider audience. A post that gets silence will be buried.²

The ranking signals

LinkedIn’s algorithm ranks content using several signals, but they are not equally weighted. The hierarchy, based on van der Blom’s research and LinkedIn’s own engineering disclosures:¹³

Signal	Weight	Why it matters
Dwell time	Highest	How long someone spends consuming your content. Now the #1 ranking factor.
Saves	Very high	1 save = 5x the reach impact of 1 like. Signals “I want to come back to this.”
Comments	High	Comments over 15 words carry 2.5x more weight than short ones. Quality over quantity.
Shares/Reposts	Medium	Signals the content is worth distributing to someone else’s network.
Reactions	Lowest	A like is the lowest-friction action and the weakest signal.

The shift is significant: LinkedIn has moved from counting engagement volume (how many likes?) to measuring engagement quality (how long did people spend? did they save it? did they have a real conversation?). This is LinkedIn’s “knowledge and advice” pivot — favouring content that teaches, informs, or sparks genuine discussion over content that simply provokes reactions.²

The core insight

LinkedIn’s algorithm is a prediction engine. It predicts whether a specific person will find your content worth their time. Every signal it tracks — dwell time, saves, comments, shares — is a proxy for “was this worth reading?” Your job is not to game the algorithm. Your job is to create content that is genuinely worth someone’s time. The algorithm will follow.

Under the hood: 360Brew

LinkedIn’s engineering team replaced traditional signal-based ranking with a system called 360Brew, which uses a dual-encoder architecture and LLM-powered semantic understanding. Rather than counting engagement metrics, it transforms user profiles, engagement history, and behavioural signals into dense vector embeddings — understanding what content means, not just how many people clicked on it.³

This means the algorithm can now match content to audiences by topic relevance, even if those people are not in your network. Your post about supply chain logistics can reach supply chain professionals three connection degrees away if the algorithm determines it is relevant to their demonstrated interests.

Part 2 — Content formats: what works and why

Not all formats are equal. Van der Blom’s data shows clear performance differences, and the reasons matter more than the numbers.¹⁴

The format hierarchy

graph TD
    subgraph Reach
        PO[Polls<br/>1.64x] --> DO[Documents<br/>1.45x]
        DO --> IM[Images<br/>1.18x]
        IM --> VI[Video<br/>1.10x]
        VI --> TX[Text-only<br/>0.88x]
    end

    style PO fill:#5cb85c,color:#fff
    style DO fill:#4a9ede,color:#fff
    style TX fill:#d9534f,color:#fff

Format	Reach multiplier	Engagement rate	Best for
Polls	1.64x	4.4%	Reach and audience research. Quick signal on what your audience cares about.
Documents/Carousels	1.45x	24.4% (highest)	Deep teaching. Swipeable format maximises dwell time.
Images	1.18x	6.1%	Visual storytelling. Photos of real people outperform stock images.
Video	1.10x	6.5%	Personality and trust. Short-form (<60 sec) dramatically outperforms long.
Text-only	0.88x	4.1%	Stories, opinions, reflections. Lowest reach but lowest production cost.

Why carousels dominate

Carousels (uploaded as PDFs or documents) are the standout format. They get 3.7x higher engagement than text-only, 278% more engagement than video, and 2.5x more shares.⁴ The reason is mechanical: carousels maximise dwell time. Users spend 15-20 seconds per carousel versus 8-10 seconds for text/image posts. Each swipe is a new engagement signal the algorithm registers.⁴

The key is completion rate. A 5-slide carousel where users view all 5 slides outperforms a 20-slide carousel where users stop at slide 8. The algorithm measures consumption, not volume.¹

Think of it like...

A carousel is a short picture book. Each slide is a page. The reader’s act of swiping tells LinkedIn “this person is still interested.” A text post is a single page — the reader either reads it or scrolls past, giving the algorithm only one data point.

Video: the nuances

Video reach has declined 36% year-over-year for company pages, but it remains powerful when done right:⁴

Short beats long. Videos under 60 seconds get 53% more engagement. The sweet spot is 30-90 seconds.
Vertical beats landscape. 84% higher engagement for vertical format. 73-80% of views are on mobile.
Subtitles are mandatory. 29% higher engagement, 32% longer retention. Most LinkedIn users scroll with sound off.
Personal beats polished. Authentic, personal-style videos get 44% more reactions than corporate motion graphics.
The first 3 seconds determine everything. 65% of user retention is decided in the opening moments.

Text-only: the underestimated workhorse

Text posts have the lowest reach multiplier but the lowest production cost. When done well, they punch above their weight:⁵

Optimal length: 800-1,000 characters for personal profiles. Long enough for dwell time, short enough to hold attention.
Readability matters: Posts at 4th-grade reading level perform 35% better. Sentences above 10th-grade level lose 35% reach.
Formatting is critical: Maximum 4 lines per paragraph. Use whitespace. The top 5% of posts use 16-20 sentences with consistent structure.

The 60-30-10 content mix

A recommended framework for sustainable content strategy:⁴

Proportion	Type	Purpose
60%	Educational carousels/documents	Teach something. Build authority. Maximise saves and shares.
30%	Thought leadership text posts	Share opinions, experiences, lessons. Build trust and personality.
10%	Engagement content (polls, questions)	Spark conversation. Learn what your audience cares about.

Part 3 — Engagement mechanics: the signals that matter

Understanding what LinkedIn counts as “engagement” — and how it weights different actions — is the difference between activity and impact.

The engagement hierarchy

graph TD
    S[Save<br/>5x a like] --> C[Comment<br/>2x a like]
    C --> SH[Share/Repost<br/>Medium weight]
    SH --> R[Reaction/Like<br/>Baseline]

    style S fill:#5cb85c,color:#fff
    style C fill:#4a9ede,color:#fff
    style R fill:#d9534f,color:#fff

Saves are the most underrated signal. One save has 5x the reach impact of one like. A saved post also correlates with 130% higher probability that the saver will follow you. When you create content people want to reference later — templates, frameworks, checklists, data — you trigger saves.¹

Comments are the engine of distribution. Posts with 12 meaningful comments exceed the reach of posts with 50 generic reactions by 3.2x. But quality matters intensely: comments over 15 words carry 2.5x more weight than “Great post!” And replies to comments (threaded conversation) achieve 2.4x higher reach than direct comments on the post. The algorithm is measuring conversation depth, not comment count.¹⁶

Generic AI comments (“Great post!”, “Thanks for sharing!”) are now filtered out 45% of the time. LinkedIn’s NLP models detect formulaic engagement and discount it.⁶

The golden hour strategy

What you do in the first 90 minutes after posting determines most of your reach:⁶

Engage before posting. Spend 15-30 minutes commenting on others’ posts before you publish. This warms the algorithm and puts your profile in people’s feeds.
Post at peak times. Tuesday through Thursday, 8-10 AM in your audience’s timezone. Posting during peak times boosts engagement by 23-47%.
Respond immediately. Posts with author responses within 30 minutes get 64% more total comments and 2.3x more views. Your replies signal active conversation.
Engage after posting. Continue commenting on others’ content for 15-30 minutes after publishing. This keeps your profile active in the feed ecosystem.

The critical rule

Never “post and ghost.” Publishing content and then disappearing is the single most common mistake on LinkedIn. The algorithm interprets your absence as a signal that the content is not sparking conversation. If you do not have time to engage for 30 minutes after posting, postpone the post to a time when you do.

Connection degree and distribution

Your content reaches different audiences at different rates:⁶

Audience	Typical reach
1st-degree connections	10-15% see your post
Followers (non-connections)	Up to 25-30% see your post
2nd-degree (friends of friends)	Reached when 1st-degree engages
3rd-degree and beyond	Reached by topic relevance (360Brew)

A counterintuitive finding: followers see your content at nearly twice the rate of connections. The follow relationship is a stronger intent signal than the connection relationship, because following is a deliberate choice while many connections are accepted passively.¹

LinkedIn’s SSI scores your activity from 0-100 across four pillars:⁷

graph TD
    SSI[Social Selling Index] --> PB[Professional Brand<br/>25 pts]
    SSI --> FP[Find the Right People<br/>25 pts]
    SSI --> EI[Engage with Insights<br/>25 pts]
    SSI --> BR[Build Relationships<br/>25 pts]

    style SSI fill:#4a9ede,color:#fff

Industry average is ~35. A score of 75+ puts you in thought leader territory. LinkedIn is gradually de-emphasising SSI in favour of AI-driven tools, but it remains the clearest diagnostic for your overall LinkedIn activity health.⁷

Part 4 — Building traction from zero

If you are starting with a small audience, the strategy is different from maintaining a large one. Here is what the data says about growing from nothing.

Posting frequency

The research converges on a clear range:⁶

Frequency	Effect
1 post/week	Baseline. Enough to stay visible but slow growth.
3-5 posts/week	Optimal. Consistent presence without oversaturation.
Daily	Aggressive but effective if quality is maintained.
2+ per day	Counterproductive. The algorithm prefers moderation; minimum 12-hour gap.

Allow 48-72 hours for algorithmic evaluation between posts. Weekend posts perform 50% lower than weekday posts.⁶

The hook

The first two lines of your post determine whether anyone reads the rest. LinkedIn truncates after ~2 lines with a “see more” link. If the hook does not compel the click, the algorithm never gets the dwell time signal it needs to distribute the post.

What works:

Hook type	Example	Why it works
Specific data	”We spent $47K testing LinkedIn strategies. Here’s what actually worked.”	Concrete, surprising, promises value
Contrarian claim	”Most LinkedIn advice is wrong. Here’s what the data says.”	Creates curiosity gap
Personal failure	”I lost my biggest client last month. Here’s what I learned.”	Vulnerability signals authenticity
Direct question	”What’s the one skill that changed your career? For me it was…”	Invites immediate participation

What does not work: “I want to share something important” or “Here are my thoughts on…” These are generic and signal nothing worth clicking for.

Niche positioning

The algorithm rewards topic consistency. When you post consistently about the same subject, LinkedIn classifies you as an expert in that niche and surfaces your content to people interested in that topic — even if they are not in your network.²

“B2B SaaS demand generation for companies under 50 employees” outperforms “marketing tips.” The more specific your niche, the more the algorithm can match you to the right audience.

Format rotation

Posting in the same format every time limits your reach. Format rotation — alternating between carousels, text, video, and polls — leads to 37% more follower growth and 28% more consistent visibility.⁶

Personal stories outperform advice

The most-shared posts on LinkedIn are not generic advice. They are specific stories: a founder sharing the exact month they almost ran out of money. A marketer breaking down a campaign that failed. Failures outperform success stories because they signal authenticity and create emotional resonance.⁶

Rule of thumb

If your post could have been written by anyone in your industry, it will perform like a generic post. If it could only have been written by you — because it draws on your specific experience, your specific data, your specific failure — it will outperform.

Part 5 — B2B, thought leadership, and the hidden buyer

LinkedIn is the dominant platform for B2B marketing and professional thought leadership. But the dynamics are more subtle than “post good content and leads will come.”

The hidden buyer

The 2025 Edelman-LinkedIn B2B Thought Leadership Impact Report introduced a crucial concept: hidden buyers — unseen stakeholders in finance, legal, compliance, procurement, or operations who hold real power over purchase decisions but are not the obvious target audience.⁸

Metric	Target buyers	Hidden buyers
Spend 1+ hr/week on thought leadership	64%	63%
Trust thought leadership more than marketing materials	—	64%
More receptive to outreach after reading thought leadership	—	95%
More likely to advocate during RFP	—	79%

The implication: your LinkedIn content is not just reaching the people you think it is reaching. The CFO who never likes your posts might still be reading them — and forming opinions about your credibility that influence purchasing decisions months later.

Personal profiles vs company pages

This is the most important strategic decision on LinkedIn:

graph LR
    subgraph Reach Comparison
        PP[Personal Profile<br/>10-15% reach] --> CP[Company Page<br/>1.6% reach]
    end

    style PP fill:#5cb85c,color:#fff
    style CP fill:#d9534f,color:#fff

	Personal profile	Company page
Organic reach	10-15% of connections, 25-30% of followers	1.6% of followers
Feed allocation	~62% of average feed	~2-5% of average feed
Engagement	8x higher than company pages	Baseline
Trust	73% of buyers trust employee-shared content more	Perceived as marketing

The data is overwhelming: people trust people more than logos. Employee posts outperform company pages by 6-8x in reach and engagement. Employee advocacy generates 561% greater reach and 7x more lead conversion than company pages.⁹

The winning strategy is hybrid: build personal brands on employee profiles while using the company page for official announcements, job postings, and paid amplification. LinkedIn’s Thought Leader Ads allow you to boost employee posts as advertisements, bridging the gap.⁹

Thought leadership that converts

The Edelman report found that 91% of hidden buyers say quality thought leadership helps them uncover challenges they had not recognised. This is the mechanism: thought leadership does not sell directly. It reframes the buyer’s understanding of their own problems, making them receptive to your solution when the time comes.⁸

Effective thought leadership on LinkedIn:

Teaches from experience, not from theory
Takes a position, not a neutral stance
Is consistent — weekly or bi-weekly, on the same topic
Addresses the hidden buyer, not just the obvious one
Shows the thinking, not just the conclusion

Part 6 — What kills your reach

Understanding what the algorithm penalises is as important as understanding what it rewards.

The penalty list

Mistake	Impact	What to do instead
External links in post body	~60% less reach	Place links in the first comment
Engagement bait (“Comment YES if you agree”)	Immediate suppression. LinkedIn’s NLP detects these patterns.	Ask genuine questions that invite specific answers
Editing after engagement starts	Reach effectively stops growing	Proofread before publishing. Minor typo fixes within 1-2 minutes are safe.
Tagging 5+ people	Flagged as spam	Tag 3 maximum, and only people likely to engage
Post and ghost	Algorithm assumes no conversation	Stay active for 30 minutes after posting
6+ hashtags	”Seriously sabotage reach”	3-5 maximum. Hashtags no longer drive discovery — they are only search keywords.
Generic AI comments	Filtered out 45% of the time	Write specific, substantive comments that reference the content
Posting 2+ times per day	Reduces reach per post	Minimum 12-hour gap between posts

The engagement pod trap

Engagement pods — groups that artificially like and comment on each other’s posts — are now actively targeted by LinkedIn. In March 2025, LinkedIn’s VP of Product Management stated the goal is to make engagement pods “entirely ineffective.”¹⁰

LinkedIn’s detection system looks for: identical engagement timing patterns, generic comment language, relationship clusters that always engage with each other, and engagement-action mismatches (someone “loves” every post from a group but never engages with similar content outside it). Penalties include shadow bans and reach collapse — one marketer dropped from 8,500 to 340 impressions overnight after pod detection.¹⁰

The link penalty workaround

Posts with a single external link lose approximately 60% of their reach. But counterintuitively, posts with 4+ links actually perform 3-5x better — because the algorithm interprets multiple links as a curated resource rather than a promotional redirect.¹¹

The standard workaround: place your outbound link in the first comment and reference it in the post body (“Link in comments”). This avoids the penalty while still making the link accessible. Some practitioners report that even this is becoming less necessary as LinkedIn’s algorithm evolves, but it remains the safest approach.¹¹

Part 7 — The full map

graph TD
    subgraph Algorithm
        FP[Feed Pipeline<br/>Quality, Golden Window,<br/>Review, Distribution]
        RS[Ranking Signals<br/>Dwell time, Saves,<br/>Comments, Shares]
    end

    subgraph Content
        CF[Content Formats<br/>Carousels, Video,<br/>Text, Polls]
        CM[Content Mix<br/>60-30-10]
    end

    subgraph Engagement
        GH[Golden Hour<br/>First 90 minutes]
        CS[Comment Strategy<br/>Quality over quantity]
        SSI[Social Selling Index]
    end

    subgraph Strategy
        NP[Niche Positioning<br/>Topic consistency]
        TL[Thought Leadership<br/>Hidden buyers]
        PP[Personal over<br/>Company pages]
    end

    FP --> RS
    RS --> CF
    CF --> CM
    CM --> GH
    GH --> CS
    CS --> NP
    NP --> TL
    TL --> PP

    style RS fill:#4a9ede,color:#fff
    style GH fill:#5cb85c,color:#fff
    style TL fill:#f0ad4e,color:#fff

What you should understand now

After reading this path, you should be able to:

Explain how LinkedIn’s four-stage feed pipeline determines which content gets distributed and why the first 90 minutes are decisive

Rank the engagement signals (dwell time, saves, comments, shares, reactions) by their algorithmic weight and explain why quality trumps volume

Choose the right content format for a specific goal (reach, engagement, trust, authority)

Apply the 60-30-10 content mix to build a sustainable posting strategy

Describe the hidden buyer phenomenon and explain why thought leadership reaches audiences you cannot see

Identify and avoid the behaviours that trigger algorithmic penalties (external links, engagement bait, editing, tagging)

Articulate why personal profiles outperform company pages and design a strategy that leverages both

Check your understanding

Test yourself (click to expand)

Explain why dwell time has become LinkedIn’s most important ranking signal. What does this tell you about what kind of content to create?

Compare the reach and engagement performance of carousels, video, and text-only posts. Why do carousels dominate despite text being easier to produce?

Design a golden hour strategy for a post you want to publish. What do you do before, during, and after the first 90 minutes?

Evaluate this claim: “A company page is the most important LinkedIn asset for B2B marketing.” Using what you have learned, argue for or against this position with data.

Interpret this scenario: you post a detailed carousel about your area of expertise. It gets 15 saves, 8 thoughtful comments, and only 23 likes. Your colleague posts a selfie with a motivational quote and gets 200 likes but 0 saves and 2 comments. Which post performed better in the algorithm’s eyes and why?

Where to go next

Path A -- Build your personal brand strategy

You understand how the platform works. Now learn the strategic layer above it: how trust builds over time, what to post at each growth stage, and how followers become clients. Building a Brand That Converts covers cadence, content mix, the trust pipeline, and the path from follower to client.

Path B -- Build your content engine

Start creating LinkedIn content using the 60-30-10 framework. Focus on one niche topic, post 3-5 times per week, and track saves and comments (not likes) as your key metrics. Explore channel-mix-strategy to understand how LinkedIn fits into a broader marketing approach.

Path B -- Understand the marketing fundamentals

LinkedIn is one channel in a larger system. To understand how it connects to customer acquisition, retention, and lifetime value, explore the e-commerce learning path and concepts like customer-acquisition-cost, conversion-rate-optimisation, and customer-journey-mapping.

Path C -- Deep dive into thought leadership

If your goal is to position yourself or your organisation as a thought leader, read the 2025 Edelman-LinkedIn B2B Thought Leadership Impact Report and explore how thought leadership connects to customer-retention and long-term brand building.

Explorer

How LinkedIn Actually Works: Algorithm, Content, and Traction

How LinkedIn Actually Works: Algorithm, Content, and Traction

Who this is for

Part 1 — The machine behind the feed

How the feed is assembled

The four-stage pipeline

The ranking signals

Under the hood: 360Brew

Part 2 — Content formats: what works and why

The format hierarchy

Why carousels dominate

Video: the nuances

Text-only: the underestimated workhorse

The 60-30-10 content mix

Part 3 — Engagement mechanics: the signals that matter

The engagement hierarchy

The golden hour strategy

Connection degree and distribution

Social Selling Index (SSI)

Part 4 — Building traction from zero

Posting frequency

The hook

Niche positioning

Format rotation

Personal stories outperform advice

Part 5 — B2B, thought leadership, and the hidden buyer

The hidden buyer

Personal profiles vs company pages

Thought leadership that converts

Part 6 — What kills your reach

The penalty list

The engagement pod trap

The link penalty workaround

Part 7 — The full map

What you should understand now

Check your understanding

Where to go next

Sources

Further reading

Footnotes

Graph View

Table of Contents

Backlinks