Remember when your LinkedIn feed was full of posts with 47 comments that all said "Great insight!" and "Thanks for sharing!"?
Yeah, LinkedIn remembers too. And they weren't fans.
In March 2025, LinkedIn rolled out what they're calling their "authenticity update"—which is corporate speak for "we're done with your engagement pod nonsense." The change was quiet. No big announcement. Just a sudden, dramatic drop in reach for anyone still playing the artificial engagement game.
I've been tracking LinkedIn algorithm changes since 2019, and this one's different. It's not a tweak. It's a fundamental shift in how the platform identifies and rewards genuine professional conversation. And if you're still using tactics from 2023, your content is basically invisible.
Let me show you what actually works now.
The Engagement Pod Massacre of 2025
Here's what happened: LinkedIn's AI got scary good at pattern recognition.
They're now tracking comment velocity, account relationships, engagement history, and even the semantic content of comments. When 15 people comment within 90 seconds of your post going live, all using generic phrases? The algorithm knows. When those same 15 accounts comment on each other's posts in the same pattern every Tuesday at 9 AM? The algorithm definitely knows.
A marketing director I know saw her average post reach drop from 8,500 impressions to 340 overnight in April. She was in three engagement pods. LinkedIn didn't suspend her account or send a warning. They just... stopped showing her content. The algorithmic death penalty.
The platform is now explicitly deprioritizing:
- Posts with suspicious early engagement patterns
- Content that gets generic comments ("Great post!" "Thanks for sharing!" "Love this!")
- Accounts that consistently engage with the same small group
- Posts with high comment counts but low dwell time
- Content that gets engagement but no profile visits or follows
Because clearly what professional networking needed was more genuine interaction and less coordinated theater. Wild concept.
What LinkedIn Actually Rewards Now
The 2025 algorithm has three primary ranking factors, and they're not what most people think.
Conversation Depth Over Engagement Volume
LinkedIn is now measuring the quality of comment threads, not just the quantity. A post with 12 substantive comments that spark actual discussion will outperform a post with 50 generic reactions.
The algorithm looks at:
- Comment length (longer generally signals genuine engagement)
- Reply depth (are people responding to each other, not just the original poster?)
- Semantic variety (diverse vocabulary and perspectives)
- Time spent reading comments (tracked through scroll behavior)
- Whether commenters actually viewed your profile or clicked your content
I tested this extensively in November. Two posts, similar topics. Post A got 43 comments, mostly short and generic. Post B got 19 comments, but they were detailed responses that sparked threaded conversations. Post B reached 3.2x more people and drove 8x more profile visits.
The math changed.
Audience Retention Signals
LinkedIn now tracks what happens after someone sees your post. Did they keep scrolling immediately? Did they read it? Did they click your profile? Did they visit your website? Did they follow you?
These downstream actions matter more than the initial engagement. A "like" from someone who immediately scrolled past is worth almost nothing. A profile visit from someone who didn't engage at all? That's valuable.
This explains why some posts with modest engagement metrics suddenly take off 48 hours later. The algorithm is testing your content with small audiences, measuring retention signals, and expanding reach based on how people actually respond—not just whether they clicked a button.
One creator I follow posted a detailed case study that got only 8 comments in the first 6 hours. Looked like a dud. Then LinkedIn's algorithm noticed that people were spending an average of 4 minutes reading it, and 23% were clicking through to his website. The post eventually reached 45,000 people. The algorithm rewarded genuine interest, not performative engagement.
Network Relevance Scoring
This one's subtle but powerful. LinkedIn is now much better at understanding the actual professional relationships in your network.
They're looking at:
- Whether you've had real conversations with commenters (DMs, previous meaningful interactions)
- If commenters work in related industries or roles
- Whether your content matches your stated expertise and job history
- If the people engaging with your content also engage with similar topics from others
Basically, if you're a B2B SaaS marketer and your engagement is coming from real estate agents and life coaches who don't otherwise interact with SaaS content, the algorithm smells something fishy.
Authenticity at scale. LinkedIn finally figured out how to measure it.
The Content Formats That Actually Perform
Format matters more now than it did 18 months ago. Here's what's working based on data from Q4 2025:
Document Carousels (But Different)
Carousels still work, but the approach changed. The algorithm now penalizes generic "10 tips" carousels with obvious stock photos and basic advice.
What performs: Detailed case studies, original data, specific frameworks with real examples. The carousels that work are dense with actual information. Think less "motivational quote on a gradient background" and more "here's the exact email sequence that generated $340K."
Average performing carousel in my testing: 2,400 impressions, 32 engagements.
High-performing carousel with specific, detailed content: 18,000 impressions, 340 engagements.
The difference? Substance. The algorithm can apparently tell when you're just filling slides with fluff.
Long-Form Text Posts
Counter to what everyone predicted, long-form posts are crushing it in 2025. I'm talking 1,200-1,500 characters of actual insight.
The key is structure. Break it into short paragraphs. Use line breaks liberally. Make it scannable. But don't dumb it down.
LinkedIn's dwell time tracking means that posts people actually read (not just scroll past) get massive reach boosts. A well-written 1,300-character post will outperform a shallow 200-character hot take every time.
This connects to broader shifts in how platforms are measuring content value—something we've seen across multiple channels, not just LinkedIn. The era of optimizing for the algorithm instead of the reader is finally ending. (Okay, it's not ending. But it's getting harder.)
Video With Actual Value
Video still gets preferential treatment, but LinkedIn got much better at distinguishing between "talking head saying generic things" and "actual valuable content that happens to be in video format."
The algorithm now measures:
- Average view duration (not just 3-second views)
- Rewatch rate (are people going back to catch something?)
- Whether viewers turn on sound (auto-play is muted, so this signals genuine interest)
- Comments that reference specific points from the video
Posting a 2-minute video of yourself sharing "3 quick tips"? Meh. Posting a 4-minute breakdown of a specific strategy with screen recordings and real results? That's what performs.
Quality finally matters more than volume. Who knew.
The Posting Strategy That Works
Forget everything you learned about optimal posting times and frequency. The 2025 algorithm cares more about consistency and relevance than timing.
Here's what's actually working:
Post 2-3 times per week, maximum. The algorithm penalizes accounts that post daily with mediocre content. Better to post twice weekly with substantial insights than daily with shallow observations. LinkedIn is rewarding depth over frequency.
Give it 48-72 hours. The new algorithm takes time to evaluate content performance. That post that looks dead after 6 hours? It might explode on day 2 if the retention signals are strong. Stop deleting posts that don't immediately perform.
Respond to every substantive comment within 2 hours. This signals to the algorithm that you're fostering genuine conversation. But here's the key: your responses need to add value. A simple "Thanks!" doesn't help. A thoughtful reply that extends the conversation? That's algorithmic gold.
Cross-pollinate with genuine relationships. The algorithm rewards when people in your network consistently engage with each other's content—but only if those relationships appear authentic. Regular DM conversations, varied engagement patterns, and relevant professional connections all signal legitimacy.
I've been testing different approaches since September. The accounts that treat LinkedIn like an actual networking platform (shocking, I know) consistently outperform those treating it like a content distribution channel.
What To Do If Your Reach Tanked
If you were in engagement pods and your reach collapsed, here's the recovery playbook:
Stop all artificial engagement immediately. Leave the pods. Stop the comment-for-comment arrangements. Cold turkey. The algorithm needs to see a pattern change.
Focus on 10-15 genuine relationships. Find people in your industry who post valuable content. Engage meaningfully with their posts. Have actual conversations. Build real professional relationships. The algorithm will notice the shift from generic engagement patterns to authentic interaction.
Post less, but better. One genuinely valuable post per week beats five mediocre ones. Give the algorithm something worth promoting.
Engage before you post. Spend 15 minutes genuinely engaging with others' content before publishing your own. This signals to the algorithm that you're an active, authentic community member—not just a broadcaster.
Recovery takes 4-6 weeks typically. I've seen several accounts rebuild their reach by going back to basics: actual valuable content, genuine engagement, real professional networking.
Revolutionary stuff, really.
The Metrics That Actually Matter Now
Stop obsessing over vanity metrics. Here's what to track in 2025:
Profile visits per post. This is the best indicator of whether your content is actually resonating. Aim for a 2-5% conversion rate (profile visits divided by impressions).
Average engagement rate on comments. Not just top-level comments, but replies within threads. This shows conversation depth.
Follower growth rate. If your content is valuable, people follow you. If your follower count is stagnant despite high engagement, something's off.
Click-through rate on shared links. When you share content, what percentage actually click? This signals genuine interest versus performative engagement.
DM conversations initiated. The best content sparks private conversations. If you're not getting DMs from your posts, you're not creating enough value or controversy.
LinkedIn's analytics dashboard added several new metrics in October 2025 specifically around engagement quality. Use them.
What's Coming Next
Based on platform signals and early testing, here's what I'm seeing for early 2026:
More emphasis on expertise signals. LinkedIn is getting better at identifying genuine subject matter experts versus people who just post about everything. Expect the algorithm to reward focused, deep expertise over broad, shallow content.
Collaborative content features. LinkedIn is testing features that make it easier for multiple people to co-create content. The algorithm will likely reward these collaborative posts, assuming the collaboration appears authentic.
Integration with LinkedIn Learning. There are signs that LinkedIn will start connecting content creators with their Learning platform, potentially giving algorithmic boosts to creators who demonstrate teaching ability.
Stricter penalties for engagement bait. Those "Agree or disagree?" posts that add nothing substantive? LinkedIn's testing ways to penalize obvious engagement bait tactics.
The trend is clear: LinkedIn wants to be a place for genuine professional development and networking, not a vanity metrics game. About time.
The Bottom Line
Engagement pods are dead. Good riddance.
What works now is what should have always worked: valuable content, genuine relationships, and actual professional expertise. LinkedIn's 2025 algorithm is sophisticated enough to measure authenticity at scale.
This is harder than gaming the system with engagement pods. It requires actually knowing your stuff and building real relationships. It takes more time and effort.
But here's the thing: the people who were creating genuine value all along just got a massive advantage. The playing field tilted toward quality. If you've been doing it right, your reach probably improved this year.
And if you've been relying on artificial tactics? Well, you've got some work to do.
The algorithm finally caught up to what real networking looks like. Adapt or become invisible.
Your move.
Top comments (0)