LinkedIn’s Proxy Problem

LinkedIn does not lack content. It lacks respect for the reader’s time

The hard part is that the people building these mechanics are not stupid. The people adopting them are not shallow. Everyone is responding to incentives that are visible on the surface and hard to ignore. If distribution is awarded to formats that force a pause, then creators will build posts that force a pause. If the platform promotes behaviors it can measure easily, then the platform will measure what is easy and call it value. That is how systems decay without anyone meaning to break them. The deeper question is not why video exists on LinkedIn. The deeper question is why the platform is willing to confuse a professional network with an entertainment feed, then act surprised when serious people start treating it as noise. The confusion is not accidental. It is the product of a measurement choice. “What would have to be true for this outcome to keep repeating.”

The Contract That Made LinkedIn Worth Opening

Most people did not come to LinkedIn to be entertained. They came because the platform solved a specific coordination problem. It let you see who a person was, what they did, who could vouch for them, and whether their ideas held up under professional scrutiny. It is hard to overstate how rare that bundle is. Identity is expensive. Reputation is expensive. A credible graph of relationships is expensive. LinkedIn made those costs smaller. The contract was never that every post would be brilliant. The contract was that the feed would contain enough signal to justify a glance. You could find a peer’s hard-earned lesson without reading a book. You could find a candidate without paying a headhunter. You could learn that a vendor had changed direction without waiting for a press release. You could find a colleague’s writing and send it to your team with a short note that carried weight because you were forwarding it, not because the platform was promoting it. That is why the platform had a natural bias toward text. Text is the most compressible professional medium. It can be scanned. It can be quoted. It can be searched. It can be saved and returned to with almost no re-entry cost. It also carries accountability. When someone writes plainly, you can test it later against reality. You can ask whether it helped, whether it held, whether it was a convenient story. Video can do those things, but only with a different infrastructure. Searchable transcripts. Indexing. Chaptering. The ability to land on the exact sentence that mattered. Without that, video is a performance format, not a working format. It asks for time up front and gives back value later, if at all. That trade does not work for the audience LinkedIn claims to serve. This is where the platform’s design decision starts to look like a breach of the original contract. When the feed becomes harder to scan, you do not merely lose minutes. You lose the habit of

opening the platform when you have a question and need an answer. Once that habit breaks, the platform can win engagement metrics and still lose relevance in the lives of the people who decide budgets, approve hires, and set direction. The irony is that professionals are not anti-media. They watch long podcasts. They sit through earnings calls. They read books. They do not hate length. They hate waste. They will give time when the format respects their intent. LinkedIn’s problem is not that it is trying to be richer. It is that it is trying to be stickier.

When Saves Replace Outcomes

Saves are a telling metric because they sit at the border between attention and intent. Saving something can mean, this is valuable. It can also mean, I do not have time right now, I will deal with it later. In a professional context, that second meaning is common. People save because their day is full, not because the content is profound. They save because the format is asking more time than they can pay in that moment. When a platform begins to treat saves as a proxy for value, it risks rewarding content that creates backlog rather than clarity. Backlog feels like productivity because it looks like a pipeline of future learning. In practice it is often a pile of unread tabs, unread emails, and unread “saved” posts that will never be reopened. A platform can increase saves while decreasing usefulness, simply by making posts harder to absorb in one pass. That is the measurement hazard. A metric can move in the right direction while the underlying product moves in the wrong one. If you optimize for the metric, you accelerate the drift. You create creators who design for saving, not for finishing. You create readers who feel behind, not equipped. You create a system where everyone is “engaging” and fewer people are learning. Saves are a symptom. Outcomes are the contract. The professional question is not whether something was saved. It is whether it changed an action. Did it tighten a decision. Did it reduce the time between a signal and a response. Did it prevent a bad hire. Did it help someone avoid an avoidable loss. Did it produce an introduction that mattered. Did it surface an idea that survived the week, not the scroll. That is why longform has always mattered on LinkedIn when it is written by people who have paid for their conclusions. Not because it is long. Because it is testable. Longform gives you enough mechanism to decide whether the author knows what they are talking about. It gives you enough context to reuse the idea inside your own constraints. It gives you enough friction to keep casual posting honest. That is precisely what a professional platform should want. But the platform’s current incentives can punish that honesty. If distribution is granted for watch time and swipe depth, then the writer who explains cleanly loses to the creator who stretches one thought into ten panels. The platform ends up rewarding packaging over substance, then calls the

result “creator success.” That is how you get all sizzle, no content, dressed up as a new media era. You can see the second-order effect in how colleagues start reposting each other. The repost becomes a hedge. If original work is punished, you borrow distribution by aligning with work that already moved. Reposting is not a sin. It is a response to a surface that no longer treats authored thinking as the highest form of value. The feed teaches people to become distributors, not writers. The platform gains movement, then loses meaning.

Video Is a Format With a Capture Cost

Video is powerful when it is the right tool. It carries voice, facial expression, and urgency. It can signal accountability, because the speaker cannot hide behind perfect prose. In moments of crisis or decisive change, video can be the cleanest medium. A leader can speak directly. A team can hear tone. The message can land with a human force that text cannot match. That is the fair case for video on LinkedIn. It is also why more executives are using it. The platform is not wrong to host it. The platform is wrong when it assumes that hosting it means privileging it. The capture cost of video is real. Recording takes setup, even when it looks casual. Watching requires presence. You cannot skim without losing the thread. You cannot quote easily without transcribing. You cannot search your own memory inside it without replaying. This is why video works best on platforms built around retrieval and long viewing sessions. They have infrastructure for search, recommendation within a topic library, and a culture that treats watching as the primary act. LinkedIn’s primary act has never been watching. It has been scanning a professional graph, then doing something. A format that raises the capture cost breaks that act. It turns the feed from a tool into a demand. Video is not bad. Unsearchable video is a tax. The tax shows up in small ways that compound. People begin to avoid opening posts because they do not want to be forced into audio. People scroll faster because they have learned that stopping has a cost. The average reader becomes less patient with text because the feed has trained them to expect hooks and edits. Meanwhile the serious reader moves their attention elsewhere. They move to podcasts they can search and bookmark. They move to newsletters that arrive with a subject line and a clear promise. They move to places where attention is treated as a resource, not as a captive. If LinkedIn wants to become a video platform, it will eventually have to build video infrastructure that respects professional intent. Searchable transcripts. The ability to copy a

sentence as text. The ability to land on the part that matters without watching the whole thing. Until then, privileging video is not product progress. It is an extraction strategy. Here is a prediction that is falsifiable and embarrassing if wrong. LinkedIn will add native, searchable transcripts for video and integrate them into on-platform search, not as a creator addon but as a default behavior. If it does not, video will remain a distribution game for creators and a time tax for operators, rather than a stable professional medium.

Carousels Are Pagination, Not Substance

Carousels are the perfect symptom of the proxy problem because they reveal the mechanics in plain view. A carousel turns one post into many swipes. It increases dwell time without increasing thought. It encourages creators to break what could have been a paragraph into a sequence that looks like substance because it takes longer to consume. It is often not substance. It is pagination. There are times when pagination is honest. A set of diagrams that genuinely require separate frames. A sequence of images where each page carries its own information. A visual story where the medium does real work. Those cases exist. They are not the norm. The norm is splitting. When the platform rewards splitting, it pushes creators to do violence to the reader. It turns the feed into a set of micro-commitments. Each swipe is a small decision the reader should not have to make. Each swipe creates a new chance to abandon, which makes the creator work harder to keep the reader, which makes the creator rely more on hooks, which makes the content less dense, which makes the whole system louder and thinner. A carousel is only valuable when the pages earn the click. The deeper cost is what this does to professional trust. Professionals are trained to notice when presentation exceeds substance. In capital allocation, that is a red flag. In operations, that is a safety hazard. In hiring, that is a reputational risk. When a platform normalizes packaging without consequence, it teaches a generation of professionals to treat packaging as competence. The damage does not stay on the platform. It travels into meetings, decks, and decisions. This is where the platform’s aspiration becomes self-defeating. LinkedIn wants to keep people on the surface longer. It does so by making the surface harder to consume. That creates a feed where people spend more time and extract less value, which trains them to distrust the feed, which causes them to spend less time, which causes the platform to push harder on mechanics, which makes the content worse. That loop is not a theory. It is the common pattern of platforms that optimize for attention metrics at the expense of intent. A professional platform should fear that loop more than it fears competition. Competition is visible. This loop is internal. It eats the platform’s purpose while it reports growth.

Two Questions That Diagnose Whether You Are Feeding the Drift

If your post performance rises while your business outcomes remain flat, what are you being paid in. Attention, or trust. When you look at who comments, are those people in positions to hire, buy, fund, promote, or partner, or are they simply other creators playing the same distribution game. If a post gets saved, do you ever see evidence that it produced action, or do you see it vanish into a backlog that no one admits exists. If your best work is longform, why is the platform teaching you to compress it into panels or clips. Is the audience truly asking for that, or is the platform forcing it because it can measure it. When you read the feed, do you feel clearer, or do you feel behind. When you close the app, do you have one idea you can use, or do you have ten ideas you have not yet processed. These are not content questions. They are governance questions. They are about whether you are letting a metric decide what kind of thinker you become in public.

Do Not Confuse a Horse With a Camel

A horse is built for speed, direction, and work. A camel is built for endurance under different conditions. Both are competent. Both are useful. Confusing them is how you end up with an animal that satisfies neither job and collapses under expectations it was never designed to meet. LinkedIn is doing that confusion in real time. It is trying to be the professional graph and the entertainment feed, the working instrument and the attention machine. The platform acts as if it can bolt the mechanics of other formats onto the professional context and get the best of both. What it gets is a feed that feels busier and teaches less. The platform’s defenders will argue that professionals are changing. That younger cohorts prefer video. That every platform must evolve. Some of that is true. Professionals do use video. They do enjoy lighter content. They do want more human texture. The error is to treat that truth as a license to rebuild the platform around the easiest-to-measure forms of consumption. A professional network has a different obligation. It deals in reputational capital. It shapes how people are perceived, hired, funded, and promoted. It is a surface where trust is transferred and sometimes destroyed. When you build that kind of surface, you do not get to pretend that attention is value. Attention is a signal. Value is what survives contact with decision. Do not confuse a horse with a camel. This is where the degradation becomes strategic. If LinkedIn continues to reward packaging over substance, it will not merely irritate heavy users. It will change the composition of who posts and who reads. Serious operators will post less often and read less often, not because they have become less social, but because the platform has become less usable. The center of gravity will

move toward creators who are optimized for format games. The platform will become louder, and the people who built its moat will spend their time elsewhere. There is a quiet tell that this is already happening. The platform itself offers a way to sort the feed by what it thinks is “top” versus what is “recent,” which is an admission that the default is not simply chronology, it is curation for engagement. Once you treat a professional feed as an engagement product, you are admitting that the platform is no longer primarily about helping people keep up with their world. It is about keeping them on the surface. A platform can live that way. Many do. A platform built on professional identity pays a heavier price for it. The end state is not that LinkedIn disappears. The end state is that it becomes optional for the people who used to treat it as a daily instrument. It becomes a place you visit when you need to broadcast, not a place you visit when you need to learn. That is the definition of a platform that has confused its own purpose. A professional network that wants to be television will eventually find itself competing with television. It will lose that fight, and it will have surrendered the one advantage it was born to defend. References This essay draws on LinkedIn’s own public engineering writing on feed ranking signals and the use of dwell time as an input to reduce low-value distribution, which clarifies how attention proxies can become design gravity, along with LinkedIn Help documentation that formalizes the difference between curated and chronological feed views. It also draws on reporting and analysis of LinkedIn’s increased emphasis on video as a growth and advertising surface, including coverage of its video monetization and brand programs and the broader move toward relevanceweighted distribution that resurfaces older posts when they continue to perform. For the economic mechanism behind proxy metrics and the limits of human consumption in informationrich environments, it leans on Herbert Simon’s work on attention scarcity and Michael Goldhaber’s early articulation of the attention economy, with additional journalistic context on executive adoption of informal video as a leadership communication tool, which serves as a partial counterexample that video can carry real value when it is accountable and intentional rather than merely rewarded by mechanics.