Feature or bug?

On platform decay.

When a product’s value proposition drifts from user incentives, the arc usually runs the same way: growth, then decay. LinkedIn is a case in point. What was once a professional rolodex now reads more like a dad joke (I’ll explain).

LinkedIn wasn’t originally built for virality. When it launched in 2003 (yes, the same year The Simple Life graced our TVs), the whole point was professional networking: you connected with people you actually knew, there were caps on how many connections you could add, and “going viral” wasn’t a function. The site was a closed social network, more like MySpace or early Facebook, where connections were real and engagement skewed one-to-one (later aided by a paid InMail subscription).

Other social media platforms like Instagram, by contrast, built and scaled on the premise of the influencer cultivating a following. Over time, LinkedIn has begun to move in that direction. It added the ability to follow people beyond your network, introduced badges like “Top Voices,” and tweaked its feed to make posts discoverable by wider audiences. In shifting away from its original value proposition, however, LinkedIn has started to show the familiar signs of platform decay.

Author Cory Doctorow calls the process of platform decay “enshittification”: the point when a platform stops serving users and starts serving itself, leading to workarounds, hacks and noise. He describes the cycle—somewhat dramatically—as follows:

First, [platforms] are good to their users; then they abuse their users to make things better for their business customers; finally, they abuse those business customers to claw back all the value for themselves. Then, they die.

You can see it on Facebook. Originally, it gave brands and creators free organic reach to audiences that had “liked” their pages. Over time, though, that visibility was throttled, forcing businesses to pay for reach they once got for free. The shift turned what had been a community-building tool into a pay-to-play channel, alienating small businesses and cluttering feeds with ads.

Twitter (now X’s) original draw was its “town square” positioning: short updates from people whose opinions you cared about. In recent years, algorithmic changes, verification-for-purchase, and heavy ad injection have warped the experience. The features that once differentiated it (real-time discovery, credibility through verified users) now feel compromised, undermining the core value proposition.

As Facebook and Twitter declined, LinkedIn benefited by comparison: it seemed like one of the last platforms untouched by toxicity, in part because it was built for professional identity rather than anonymous posting.

A key sign of decay is when you start to notice compensatory behavior: the workarounds users invent to keep their content visible and incentives aligned with the platform. Sometimes, these even form the basis of new companies. Instagram never allowed hyperlinking individual posts, which gave rise to “link in bio” workarounds and entire businesses like Linktree. For LinkedIn, we now have companies like Scripe, an AI tool specifically for LinkedIn.

Source: Scripe

Carmen Vicente, a social media manager, recently mapped LinkedIn content along two axes: contrived–authentic, and respected–cringe. “Thrilled to announce” posts and engagement pods land squarely in contrived-cringe; meanwhile, creators who subvert those clichés fall in authentic-respected.

The grid is funny, but it also marks out the core problem: the platform incentivizes content that looks like participation but isn’t. The result is a feed that too often lands in the worst quadrant: not professional rolodex, not authentic conversation, but something closer to a dad joke—desperation dressed up in sincerity (see: what witnessing my wife give birth taught me about B2B sales).

Source: Carmen Vicente

Users are noticing, and coming up with their own workarounds (scaffolding to hold up the decay, as it were). And once you spot them, they’re everywhere. The experience becomes flattened; the scaffolding a particularly obnoxious symptom. These include:

• Links in comments. Like every platform, LinkedIn is incentivized to keep you on-site. Sending users off-platform is bad for its metrics, so posts with external links are throttled and shown to fewer people. The workaround became standard: put the link in the comments. That way the post performs better, even if it means burying the actual resource and creating a frustrating extra step for users (both the posters and the readers).

• Comment-for-playbook. LinkedIn rewards engagement. A post with hundreds of comments signals that engagement, even if every comment is just “GROWTH” or “INTERESTED.” That incentive has spawned the tactic of offering a “free playbook” if people comment, then delivering the link later via DM. The algorithm could prune this by prioritizing unique comments, but as far as I can tell, it hasn’t.

• AI slop. The platform rewards consistent content that is quick to skim. AI makes it easier to churn out posts heavy on rhetorical questions, bulleted lists (emojis-as-bullets for extra cringe), and em-dashes. It isn’t that LinkedIn built the slop, but its incentives shape it.

To test this last point out, I ran a draft of this Substack post through the AI-for-LinkedIn tool Scripe to see what it would suggest. The results weren’t bad, but they felt a bit formulaic.

Source: Scripe

It isn’t just about low-effort AI content: the algorithm shapes how even thoughtful opinions get expressed. Marketing strategist Christina Le writes:

You know how I know opinions online often suck? Because it’s too easy to change them. Take LinkedIn …You might have a nuanced "hot" take, but you shave off the edges because the algorithm rewards snappier, b&w statements. This leaves us with two poles: contrarians whose entire identity is carved out of disagreement (“I’m the one who calls out the BS”), or bandwagoners who echo the safe takes for easy approval.

Contrary to everything above, the purpose of this post isn’t to slam LinkedIn. If I wanted that kind of hot take, I’d strip out the nuance and post it on LinkedIn (kidding!) The point is to show how platform decay creeps in, and think about what that means when engaging with a network that founders, marketers, and creators can’t really afford to ignore.

These aren’t unsolvable problems. There are awesome creators out there like Jillian Richardson, whose whole platform is about zigging instead of zagging (she’s even launched a newsletter and course, Make LinkedIn Weird Again.)

LinkedIn could also fine-tune its algorithm to reward substance over scale: prioritize comments with unique content, differentiate high-quality link-outs from mass-syndicated reposts, and flag posting patterns that suggest AI filler. (Personally, I’ve been trying to tone down my own robotic, drone-y comments like “agree!” or “congrats!”)

But all of this requires intention. And the gravitational pull of enshittification is toward growth metrics, not quality: for platforms, and for the people who depend on them.