Somewhere on a train this morning, one of your readers took out her phone. She did what everyone does before their stop arrives. She opened a feed and started scrolling. A clip from a press conference. A stranger’s furious opinion about the clip. A recipe she’ll never cook. An argument. An ad dressed up as an argument (sighs). Forty minutes passed, and she couldn’t tell you a single thing she chose to see. This is what social media looks like. It was supposed to free us from the “biased editorial choice” of the legacy media. Indeed, it did. To make us slaves to “algorithms”.
Then your newsletter arrived in her inbox.
Same phone. Same thumb. A completely different transaction. She had read fifty things this week, decided that five of them deserved her attention, and signed her name to the decision.
Most of the newsletter vs feed conversation gets stuck on format. Email against app. Text against video. Inbox against timeline. The real difference sits a layer deeper, in how the contents get picked. A feed is computed. A newsletter is chosen. And the chooser is the entire product.
What’s the Difference Between a Newsletter and a Feed?
Strip away the interface and a feed is a prediction engine pointed at one goal. For every piece of content in its inventory, it estimates the probability that this specific person will keep looking. The estimate wins or loses thousands of times a day and the winners fill the screen. The post your reader just saw was selected by a model betting she would linger on it. Whether it was true, useful, or worth her morning never entered the calculation.
That engine now carries most of how people meet the world. The Reuters Institute Digital News Report 2025 found that 61% of online news consumption comes through platforms like Facebook, YouTube, X, Instagram, and TikTok, with 29% coming through news websites and apps.
Let that one land before you keep reading. Did I say social media content consumers are enslaved by the algorithms? Yes, I did.
The time inside the engine tells the same story. DataReportal’s Digital 2025 analysis of GWI survey data puts the typical internet user at 2 hours and 21 minutes on social platforms every single day. Two hours of selections, with every one of them made by a system that has never met her and made for a goal she never agreed to.
The design is honest about its intentions even when the marketing is shy about them. Aza Raskin, the designer most often credited with popularizing the infinite scroll, told the BBC that engagement driven interfaces work like “behavioral cocaine” sprinkled across the screen, and he has spoken openly about the guilt of watching what the pattern did to people’s time. The scroll has no bottom because a bottom would be a decision point, and decision points are where readers leave.
Your newsletter is built out of decision points. It starts. It ends. It contains five things because a person decided the sixth thing did not belong. Every one of those properties is the opposite of an engagement estimate. Every one of them is authored.
A feed answers the question of what will hold your reader. A newsletter answers the question of what will serve her. Only one of those questions has her interest at its center.
The distinction was visible long before either object existed. In 1971, the economist Herbert Simon, who would go on to win a Nobel Prize, gave a talk called Designing Organizations for an Information Rich World. His argument has aged into prophecy.
Simon observed that information consumes the attention of the people receiving it, which means “a wealth of information creates a poverty of attention.” The systems people would actually need, he argued, were the ones that excel at filtering, because the scarce resource was attention all along.
Fifty years later, both halves of his prediction arrived. The wealth of information became the feed. The filter became the most valuable thing a reader can hold. The only open question is who operates that filter for her. An engagement model owned by a platform, or a person whose judgment she decided to trust.
Why Are Newsletters Better Than Social Media Feeds?
Better is doing specific work on that question. On reach, the feed wins. On speed, the feed wins. The case for the newsletter lives in what the feed quietly costs its readers, and that bill has been arriving for years.
The first cost is sameness. The writer Kyle Chayka spent a decade documenting what algorithmic recommendations do to culture, and his 2024 book Filterworld names the result plainly. Recommendation systems promise personalization but deliver homogenization because every feed optimizes toward the same engagement math and that math keeps surfacing the same things. Chayka calls the accompanying feeling algorithmic anxiety, the low hum of wondering whether you genuinely like what you like or the system trained you into it.
Harsh? Open your own feed tonight and count how many posts you would have chosen on purpose. Yeah, you already know the answer.
The second cost is trust. In the same Reuters report, 58% of people said telling the truth from falsehood online is getting harder, and online influencers and personalities tied to national politicians were the most frequently named threats for spreading misleading information, at 47% each. The feed delivers more content from more voices than any medium in history, and its readers trust the delivery less than they have trusted anything.
Sameness plus suspicion is the environment your newsletter lands in. It also explains the strange durability of a format older than the web. An issue in the inbox arrives carrying the two things the feed structurally cannot offer. A finite shape, and a name attached to the choices.
I wrote a few weeks ago about the contract your readers signed when they subscribed, the expectation that a specific voice keeps showing up in their inbox. The curatorial side of that contract gets less attention than the writing side, and it may carry just as much of the relationship. Your subscribers handed you their attention on the understanding that a person, the same person, would keep deciding what earns it.
A feed shows your reader what everyone is looking at. A newsletter shows her what you looked at and kept. Only the second one can carry a relationship.
Picture the two side by side one more time. An algorithmic feed is a waiter who brings out whatever the kitchen most needs to sell tonight. A newsletter is a friend who has eaten everywhere in town and books you a table.
Is Content Curation a Skill?
And now, the part nobody tells newsletter creators. The “deciding” is the craft.
You’d be in good company if that sentence feels strange. Almost every creator we talk to treats their selection work as overhead, the chore standing between them and the real work of writing. The word curation still smells like museums, so the work goes uncounted and unclaimed.
The publisher Michael Bhaskar wrote a whole book making the opposite case. In Curation: The Power of Selection in a World of Excess, he argues that once any market crosses into overabundance, value migrates from making things to selecting and arranging them. Museums learned this first. Record shops learned it next. The entire profession of the DJ is the same lesson set to music. The newsletter is where the lesson arrived in publishing, and most of the people doing the work never got the memo that it was valuable.
The behavioral evidence is older than the creator economy. In 2000, psychologists Sheena Iyengar and Mark Lepper set up a jam tasting table in a California grocery store, alternating between 24 varieties and 6. The big table drew the bigger crowd. The small table sold the jam. Shoppers who stopped at the limited display were roughly ten times more likely to buy than shoppers facing the full assortment.
The study became famous as the paradox of choice. For a newsletter creator, it reads better as a job description. The 24 jam table is the feed, infinite, glittering, and paralyzing. Your issue is the 6 jam table. The value you create lives in the eighteen jars your reader never has to consider.
Think about what your reader actually receives when your issue lands. She receives the five stories, yes. Underneath them, she receives something she would never name out loud: relief. The week produced thousands of things she could’ve read, and a person she trusts already did the reading, the comparing, and the discarding on her behalf. The issue is the visible deliverable. The relief is the real product.
Curation is the art of saying no on your reader’s behalf. Every no is invisible, and your subscribers are paying you for all of them.
Cagri mapped the operational anatomy of this work earlier in the series, the discovery, filtering, and scoring chain that eats creator hours every week. The mechanics are his lane. The meaning is what I want you to take from this one. The filtering chain feels like drudgery because the tools have always been bad at it, and the badness of the tools has convinced a generation of creators that their most distinctive work is a chore.
Marketers have started noticing the same truth from the reader’s side. Acrelia’s analysis of curated newsletters points out that manual selection reads as differentiation precisely in the sectors where algorithms dominate, because a visible human choice humanizes the message. The scarce thing now is evidence that a person was involved. Your selections are that evidence, delivered weekly.
Why Do Readers Trust a Human Chooser Over an Algorithm?
Because people sort recommendations into two categories, and your newsletter lives in the one where machines lose.
In 2020, marketing researchers Chiara Longoni and Luca Cian published a series of experiments in the Journal of Marketing on what they called the word of machine effect. The pattern held across study after study. When people weigh a recommendation on functional grounds, things like efficiency, accuracy, and practicality, they will often prefer the algorithm. When the decision involves taste, feeling, and experience, what the researchers call the hedonic realm, people resist the machine and reach for a human recommender. The belief runs deep. People simply consider humans more competent at judging what another human will love. Also, humans have evolved in this way: they always think that only another human can understand their feelings.
Map your newsletter onto that finding. A reader hunting for the cheapest flight wants the algorithm. A reader deciding how to spend twenty quiet minutes with a coffee is making a taste decision, a feeling decision, a trust decision. She wants a person on the other end of it.
You have run this experiment on yourself without noticing. A streaming app has recommended you a hundred films, and you can barely remember accepting one. A friend with good taste recommends a single film over dinner, and you watch it that weekend. The friend wins because her recommendation carries a judgment you can vouch for, made by someone who knows both the film and you.
Remember the jam table? This is the same preference, measured in trust.
The finding also explains why generic AI curation keeps disappointing the people who try it. A model trained on everyone’s engagement can tell you what performs. Your taste sits somewhere it has never been, in the pattern of what you have kept and what you have turned down across every issue you have ever published. I wrote about the writing side of this gap, the specific things a generic model erases from a draft. The curatorial version is quieter and arrives earlier, at the moment of selection, before a single sentence exists.
This is the line HeyNews was drawn along. The system reads your archive to learn which kinds of stories you have historically chosen, scores incoming candidates against that pattern, and surfaces them. Smart Select proposes a lineup. The choice stays with you on every issue, because the choice is the product your readers subscribed to.
A tool can shorten the path to your decision. The moment a tool starts making the decision, your newsletter starts becoming a feed with your name on it.
How Do You Find Your Curatorial Voice? Run the Omission Ledger
You’ve been treating your rejections as waste. They’re the most original work you do all week, and you throw the evidence away every Sunday.
Your published issues show readers what you chose. Your taste, the actual fingerprint of your editorial identity, lives just as much in everything you turned down. No analytics dashboard captures it. No tool logs it. Almost no creator has ever seen their own filter written down, which is exactly why this exercise exists.
The Omission Ledger makes the filter visible. One issue, one notebook page or spreadsheet, three steps.
Step 1: The Candidate Log. During your next production session, write down every story you seriously considered for the issue. Every article you opened with intent, every link you saved, every item that survived your first glance. Most weekly creators land between 10 and 25 candidates. Log them as you go. Reconstructing the list from memory misses the quick rejections, and the quick rejections are data too.
Step 2: The Reason Column. For every candidate that missed the issue, write one honest line about why. Skip the polite version. “Everyone already covered it.” “Right topic, wrong week.” “Smart but boring.” “It contradicts what I argued in March, and I have yet to change my mind.” “My readers would skim it.” The reasons will feel obvious as you write them down. That obviousness is the point. You are transcribing judgments you normally make in under five seconds and never record anywhere.
Step 3: The Pattern Read. At the end of the week, read only the reason column, top to bottom. Group the reasons that repeat. Most creators find three to five recurring filters they could never have named in advance. A freshness filter. A fit filter tuned to a weirdly specific picture of their reader. A redundancy filter tracks what the audience has already seen elsewhere. And maybe a courage filter, the stories rejected for being uncomfortable, which is worth noticing for an entirely different reason.
That reason column is your curatorial voice, written in your own hand. A feed has no reason column. An engagement model rejects nothing for being beneath your readers, off your beat, or beautifully made but wrong for this particular Tuesday. The ledger is one page of proof that your newsletter is authored at the level of selection, before the writing even begins.
The first run tends to produce one specific surprise. Most creators expect their filters to be about quality, and they discover the filters are mostly about their readers. Reason after reason turns out to reference a particular person’s needs, boredom thresholds, and prior knowledge. That discovery reframes the whole job. You’ve been thinking of yourself as someone who finds good stories. The ledger shows you have been doing something harder and rarer, finding the right ones, for a specific room of people you have come to know.
Run it once and you’ll start noticing the filters operating in real time. Run it quarterly and you can watch your editorial identity sharpen, the same way your archive tracks your writing voice across the years.
Cagri argued this week that everything recurring in your production belongs inside a system, in his piece on why one click should be the standard for a weekly issue. He’s right about the machinery. The ledger is my answer to the question his post leaves open. The reasons in that column are the one part of your newsletter that should stay out of every system forever, because they are the difference this whole post is about.
It All Comes Down To…
- A feed selects by predicting engagement, while a newsletter selects by human judgment with a name attached. That single structural difference, computed against chosen, explains why the two formats build completely different relationships with the same reader.
- The feed environment produces sameness and suspicion at scale. Algorithmic recommendation homogenizes what everyone sees, and 58% of people now say the truth is getting harder to discern online. A finite, signed, human selection answers both problems at once.
- Curation is a craft with evidence behind it. Value migrates to selection in any market of overabundance, and the jam study showed shoppers were roughly ten times more likely to act when a chooser narrowed 24 options down to 6. Your issue is the small table.
- Readers structurally prefer human choosers for taste decisions. The word of machine research shows people resist algorithmic recommendations in the hedonic realm, which is exactly the realm a newsletter occupies.
- You can see your own curatorial voice this week by running the Omission Ledger: log every candidate story, write one honest line for each rejection, and read the reason column at the end. The page it produces is a portrait of the judgment your readers actually subscribed to.
Go back to the reader on the train. Tonight she’ll scroll again, because everyone does, and the feed will hand her two more hours of selections that served someone else’s goal. Then, on the morning your issue arrives, she will open something a person chose for her, ended on purpose, and signed.
She can get content anywhere. The feed holds more of it than you’ll ever produce and moves faster than you’ll ever publish. What she can only get from you is a chooser. Someone whose quiet nose she trusts. Someone whose taste she borrowed on purpose, the day she typed her email address into your form.
That’s the difference between a newsletter and a feed. It rides underneath the format, in the judgment that picked five things and let the rest go. The difference is you.
See how it looks like when the choosing stays with you: heynews.co
