Why listen to me
I'm not here to criticise LinkedIn. I'm here because I need it to be better.
Before I get into what's broken, let me tell you who I actually am on this platform, because that context matters more than anything else I'm about to say.
I open LinkedIn every single day. Not out of habit. Not because my job requires it. I open it because I genuinely believe it is the most powerful professional tool available to me right now. I scroll to stay current in HR and product management. I post because I'm actively building a presence I care about. I apply for roles because I'm mid-transition into a PM career and LinkedIn is where that transition is happening. I connect with people because I've seen firsthand what one right relationship can do for a career.

I'm not a casual user with one complaint. I'm a daily user who has experienced this platform from multiple directions, and that's exactly why I think you should hear me out.
Oma, HR Professional, MBA, SHRM-CP, CPHR Candidate, PM in transition. LinkedIn user, every single day.
How I actually use LinkedIn
Feed reader. I scroll daily to stay current across HR and PM. What's in my feed genuinely shapes how I think about my work.
Content creator. I post to build my brand through a career transition. Every post I write is an investment in my LinkedIn presence, which makes the feed quality problem personal.
Job seeker. I'm actively exploring PM roles. I apply, I check listings, and I feel the silence on the other side in real time.
Former recruiter. I've used LinkedIn to hire. I know what it looks like from both sides of the job marketplace, and that background is part of why the ghost job problem bothers me as much as it does.
What I'll defend about LinkedIn, even on a bad day
The network is genuinely irreplaceable. There's nowhere else on earth where I can reach a VP of Product at a company I admire, connect with an HR leader in a different country, or find a PM community that understands what I'm building, all in one place. That's not a small thing. It's the reason I keep showing up even when the platform frustrates me. And honestly, that loyalty is what gives me the right to be honest about what needs to change.
Why you should hear this particular person out
You should listen to me not because I'm the loudest voice in the room, but because I'm still in the room. I chose to stay. I'm telling you why, and what would make me glad I did.
1 of 7
Problem 1, the feed
LinkedIn used to be where I went to learn something. Now I go to wade through things.
That shift sounds small. It isn't. It's the difference between a platform I trust and one I tolerate, and it happened gradually and then all at once.
What the feed used to be, and what changed
"The LinkedIn feed used to feel like a curated room of smart people talking about things that mattered to my career. I'd open it with actual intent, wanting to know what was shifting in HR, what product managers were thinking about. I found things there that changed how I approached my work."
In January 2026, LinkedIn quietly rolled out what analysts are calling the "Depth Score" system, measuring reading time, carousel clicks, and video completions instead of reactions. This was not communicated to users or creators. Almost overnight, content engineered for engagement started outranking content created for insight. Three-week-old posts began appearing at the top of feeds as if they were new. AI-generated carousels with titles like "10 lessons I learned the hard way," applicable to any industry, in any country, at any time, started filling the space where actual professional conversation used to live.
Now I open the app and I find posts from three weeks ago sitting at the top of my feed like they just happened. I click "not interested" and the same category of content shows up the next day. I scroll past AI-generated motivational carousels that could have been written about any profession, by anyone, at any point in time. The feed doesn't feel like it knows me anymore. It feels like it's been optimised for something I'm not actually the target of.
The feeling most mornings that something important has quietly been taken away.
Numbers that confirm what I feel
Organic reach for creators, 2025down 50%
Overall platform engagementdown 25%
Follower growth ratedown 59%
Source: Algorithm InSights 2025, Richard van der Blom
What LinkedIn gets right here, and why the gap hurts more because of it
LinkedIn's feed has always had one advantage over Twitter and Instagram: professional intent. People come here to learn, not just to scroll. That intent is the most valuable thing the feed ever offered. The algorithm change didn't just reduce reach. It eroded the very thing that made the feed worth opening in the first place. When you optimise a professional network for engagement, you keep the network and lose the professional.
Why this matters beyond my personal frustration
The feed is the front door. When it loses my trust, I stop coming with curiosity and start coming only out of necessity, job searching, responding to messages. That shift is invisible in daily active user numbers until a competitor shows up that gives serious professionals somewhere better to go. Loyalty and dependency look identical in a dashboard. They behave very differently the moment an alternative arrives.
2 of 7
Problem 2, the job marketplace
I applied with real hope. The silence on the other side taught me something I didn't want to know.
As someone navigating a deliberate career transition, the job marketplace isn't an abstract feature. It's where I've put real time, real energy, and real hope. What I found there changed how I use this platform.
What the marketplace promised, and where it started to break down
"LinkedIn's job marketplace made a promise that felt real: apply here and you're applying to something legitimate. The platform's professional reputation transferred to every listing it hosted. If it was on LinkedIn, the thinking went, it had to be real."
Ghost jobs, listings posted with no real hiring intent, left open for months to collect applicant data or make a company look active, became a systemic problem rather than an edge case. AI-powered fraud listings appeared alongside them, indistinguishable from real opportunities until it was too late. LinkedIn's own internal research eventually confirmed that 1 in 3 job seekers had experienced fake or inactive listings. The marketplace that built its reputation on professional trust started hosting the very thing that erodes it.
I've spent real hours on applications that went nowhere. Not rejected, just nowhere. No response at all, not even an automated one. I've tailored cover letters. I've researched companies. I've clicked Easy Apply with genuine hope. Coming from an HR background, I know what job searching actually costs a person before they even get to an outcome: the emotional labour, the vulnerability, the quiet hope you carry into every application. The idea that some of those listings were never real isn't just frustrating. It's a specific kind of betrayal from a platform I trusted.
Someone who has been on both sides of the hiring table and understands exactly what this costs people.
The scale of the problem
US listings with ghost job signs27% or more
ResumeUp.AI, September 2025
Job seekers who hit fake or inactive listings1 in 3
LinkedIn internal research, confirmed 2025
Rise in job scams, first half of 2025up 19%
FTC 2025, $300M lost by Americans
What LinkedIn gets right here, and why this gap is so costly
LinkedIn's job marketplace has a genuine strength that Indeed and Glassdoor can't match: professional context. When I see a listing on LinkedIn, I can see who posted it, their tenure at the company, their connections, their recent activity. That context is a real trust signal. The tragedy is that this strength, professional verification through social proof, is exactly what's being eroded by the ghost job crisis. The very thing that makes LinkedIn's marketplace unique is the thing currently being undermined.
3 of 7
Solutions
I'm not here to just complain. Here are specific ideas, with honest trade-offs attached.
Each recommendation names what it solves, what it costs LinkedIn to implement, and what the platform would have to accept in the short term. Real product thinking doesn't pretend fixes are free.
Feed quality
Job marketplace
Recommendation 1
Give users actual control over their feed, not just a "not interested" button that doesn't work
1
A feed preferences panel that actually means something
Let users select the topics they care about, not just industries but specific themes like people analytics or product strategy. Let them choose content types: posts, long-form articles, polls, events. Let them set the balance between content from people they follow versus company pages. Spotify lets me tune my Discover Weekly. LinkedIn should let me tune my feed. The data already exists because I've been clicking "not interested" for months. Use it.
2
Show the date on every post, prominently, and let me filter by recency
If the algorithm is going to surface a three-week-old post, be honest about it. Put the date somewhere visible, not buried in metadata. Give me a "last 48 hours" toggle. Let me decide whether I want to see older content today.
3
Tell creators what the algorithm is measuring
When the Depth Score system launched in January 2026, creators saw their reach drop significantly with no explanation from LinkedIn. That's not fair to people who've spent years building an audience here. Tell them what signals matter. Give them a simple view of how their content is performing on depth metrics. Treat them like partners in a platform you both depend on.
The trade-off I'm asking LinkedIn to accept
User-controlled feeds will likely reduce raw engagement metrics in the short term, because low-quality high-engagement content gets filtered out. I'm asking LinkedIn to accept that decline in exchange for long-term trust and retention. The users most at risk of churning right now, serious professionals, daily creators, active networkers, are exactly the ones whose feeds feel most broken. Fix the feed for them and you keep the users worth keeping.
Recommendation 2
Make listings earn their place on the platform before they go live, not after fraud has already happened
1
A tiered trust badge system on every listing
Three badges: Verified Company, meaning the employer domain has been confirmed. Verified Hiring Manager, meaning the poster's identity has been confirmed against the company page. Active Listing, meaning the recruiter has engaged with this specific listing in the past seven days. No badge means the listing is clearly flagged as unverified. Job seekers can filter to verified listings only. This doesn't eliminate all fraud. It makes bad actors work harder and gives good-faith users a meaningful signal to go on.
2
Automatic listing expiry after 30 days of recruiter inactivity
Any listing open for 30 days with zero recruiter activity, no profile views from the listing, no application reviews, gets automatically archived and removed from search results. To relist, the recruiter has to actively confirm the role is still open. This clears ghost jobs from the index without requiring LinkedIn to investigate each one individually. It scales.
3
One-tap fraud reporting with a 24-hour review commitment
A visible "Report this listing" button on every job post, not buried in a dropdown, that immediately suspends the listing pending review. LinkedIn commits to reviewing within 24 hours. The person who reported it gets a status update. This is table stakes for a platform that handles people's professional livelihoods.
The trade-off I'm asking LinkedIn to accept
Verification requirements and automatic expiry will reduce the number of active listings in the short term. Some employers will find the added friction annoying. That's a real cost, and I'm not pretending it isn't. But I'd ask LinkedIn to weigh that cost against this: 1 in 3 of its users has already been burned by a fake or inactive listing. Listing volume can recover. Trust, once it keeps eroding, may not.
4 of 7
Validation
My experience is the starting point. The data is what makes it an argument.
I want to be clear about something: my frustration isn't the argument. It's just what pointed me toward the problem. Here's the evidence that confirms these aren't just my bad days on the platform.
Feed
Job marketplace
Other voices
Organic reach drop
-50%
Algorithm InSights 2025
Engagement drop
-25%
Same source
Follower growth drop
-59%
Algorithm InSights 2025
Depth Score deployed
Jan 2026
No announcement made
What these numbers mean beyond the headline figures
These metrics don't just reflect pain for creators. They signal declining value for readers too. When content quality drops, both sides of the feed equation lose out. Creators stop investing in quality posts because the reach isn't there. Readers stop finding valuable content because quality creators have disengaged. It's a flywheel, except it's spinning in the wrong direction. The algorithm change that was supposed to reward quality has, so far, mostly rewarded content engineered for engagement instead.
Ghost listings, US
27%+
ResumeUp.AI, Sept 2025
Users hit by fakes
1 in 3
LinkedIn internal research
Job scam rise H1 2025
+19%
FTC 2025
Lost to job fraud
$300M
FTC 2025, first half only
LinkedIn's own acknowledgement of the problem
LinkedIn confirmed ghost job complaints increased in 2025. Their own internal research found 1 in 3 job seekers had experienced fake or inactive listings. LinkedIn's VP of Product acknowledged the issue publicly in January 2026 and outlined detection measures being added. The platform knows. The question is whether the pace of the response matches the pace of the problem, and based on current data, it doesn't yet.
Source: Social Media Today, LinkedIn VP of Product Oscar Rodriguez, January 2026
"Why is my feed full of posts that are 2 or 3 weeks old? Whenever I click not interested, what even is the subject? There is no way to fine-tune the types of items I get notified about. It doesn't matter how many times I click, I still get the same noise."
David Pereira, LinkedIn, February 2025. That post received 229 comments, the overwhelming majority agreeing.
"Navigation changed, hiding frequently used actions behind new menus, increasing taps to complete common tasks. Visual changes reduce scannability. The app feels cluttered and harder to use than it was."
Aggregated from Quora analysis of LinkedIn UX feedback, 2025
"It is sadly the last island where we can have some semblance of professional conversations, but the platform itself is also to blame."
David Pereira, LinkedIn, February 2025
What "last island" tells us about where LinkedIn actually stands
Users aren't leaving because they found something better. They're staying despite the frustrations because there's nowhere else to go. That's not a sign of a healthy product. It's a sign of a product coasting on its network moat while slowly spending down its user trust. That moat won't last forever, and the window to fix this on LinkedIn's own terms is open right now. It won't always be.
5 of 7
Performance metrics
How we'd actually know if any of this is working.
Shipping features isn't the same as solving problems. Here's how I'd measure whether these changes delivered, including the short-term costs I'd expect to see before the improvements show up.
Feed metrics
Job marketplace metrics
What gets worse first
North star: unprompted return rate
15% increase in 6 months
The percentage of daily sessions that start without a notification trigger, meaning users opened LinkedIn because they actually wanted to, not because something pushed them. This is the truest measure of whether people find the feed valuable on its own.
Why this one: engagement bait inflates total session counts. This metric filters that out and measures genuine pull, which is what matters for long-term retention.
Feed relevance score
65% positive responses within 3 months
A weekly in-app prompt to a random sample of users: "Was your feed relevant today?" Measured as a percentage of positive responses. Currently estimated below 50% based on publicly available sentiment data.
Why this one: it's a direct user signal that's harder to game than engagement metrics, and it captures the subjective experience that actually determines whether people come back.
Creator retention at 6 months
10% improvement over current cohort
The percentage of users posting at least twice a month who remain active six months after feed controls launch, compared to the same cohort six months prior. Creator health leads platform health. A feed without quality creators eventually has nothing worth reading.
Why this one: creators are the leading indicator. Their departure signals feed decline months before reader churn shows up in aggregate data.
North star: verified listing rate
60% at 6 months, 80% at 12 months
The percentage of active job listings carrying at least one active verification badge. A verification system with 20% adoption doesn't solve the trust problem. This tracks actual penetration, not just whether the feature exists.
Why this one: the only way to know if the trust system is working is to measure how much of the marketplace it actually covers.
Ghost listing rate
Under 10% within 12 months
The percentage of listings open for 30-plus days with zero recruiter activity. Currently above 27%. Automatic expiry should drive this down significantly within the first year of implementation.
Why this one: this is the core problem metric. Every percentage point down represents real job seekers who aren't wasting real hope on listings that were never going anywhere.
Job seeker NPS, isolated from overall platform NPS
15 point improvement within 12 months
"Would you recommend using LinkedIn to find a job?" Measured separately from overall LinkedIn NPS. The job marketplace is a distinct product and it deserves its own health signal, not one that gets masked by features unrelated to job searching.
Why this one: overall NPS can hide marketplace failures. Isolating this removes that cover and holds the marketplace accountable on its own terms.
Metrics that will get worse before they get better, and why that's okay
If these changes are done properly, some numbers will decline in the short term. LinkedIn should expect this and plan for it, rather than using it as a reason not to act.
Total job listings
Will drop as ghost jobs expire automatically. That's the right outcome. Fewer real listings is categorically better than more fake ones. Measure quality, not volume.
Total feed impressions
May drop as AI content gets filtered and old posts stop resurfacing. Watch the unprompted return rate alongside this. If that rises while impressions fall, the feed is getting healthier, not weaker.
Employer posting rate
May dip initially as verification adds friction. A 30-day grace period and a simplified verification flow can manage this. Monitor employer NPS separately from job seeker NPS so neither masks the other.
6 of 7
Closing
I'm still here. That's the whole point.
After everything I've described, the feed that frustrates me, the listings that disappeared into silence, I still open LinkedIn every morning. That loyalty deserves to be met with honesty. This is mine.
I'm a Nigerian-Canadian HR professional building a career in product management. I use LinkedIn to stay current, to post, to search, to connect, and to show up as someone who takes her professional life seriously. I'm building two apps. I recently completed my MBA. I'm mid-transition into a career that LinkedIn's own network has helped me pursue.

LinkedIn has given me real things. The network. The visibility. Access to conversations I wouldn't have found anywhere else. I'm not here to be angry about that.

I'm here because I think this product can be so much better than it currently is, and I care enough about it to say so clearly, with evidence, and with specific ideas attached. The best feedback comes from someone who knows you well enough to be honest with you. I know LinkedIn well enough. This is me being honest.
Oma, submitted to the LinkedIn product team with genuine belief that it matters.
What I'm asking LinkedIn to do
Give users real feed control rather than algorithmic guessing. Make job listings earn their place with verification before they go live. Name the trade-offs publicly rather than pretending changes are costless. Measure job marketplace health separately from overall platform health. Accept short-term metric pain in exchange for long-term trust.
What I'm not asking for
A perfect product, that's not the standard anyone should be held to. Fast fixes for problems that took years to develop. Changes that ignore the real complexity of building at LinkedIn's scale. Anything that compromises the network, which remains genuinely irreplaceable and worth protecting.
7 of 7