Sliding the printed spreadsheet across the mahogany table was a mistake I felt in my marrow before the paper even stopped moving. The editor-in-chief didn’t look at the rows of descending numbers or the red-shaded cells indicating a 24 percent drop in engagement from the previous quarter.
He looked at me with the kind of weary disappointment usually reserved for people who try to explain the plot of a dream they had. To him, these numbers were a foreign language spoken by someone he didn’t particularly like. To the head of audience sitting across from him, those same numbers were the only tether to reality left in a collapsing building.
The red-shaded cells indicated a quarterly decline that sparked the civil war in the boardroom.
The Vocabulary of Collapse
They spent the next talking past each other. It was a ritual dance I had seen in at least 14 different newsrooms over the . The editor spoke of “the feel of the front page” and the “moral weight” of a lead story.
The audience strategist spoke of “funnel optimization” and “scroll depth.” It was like watching a sculptor and an architect argue over a house, where one cared only about the texture of the stone and the other cared only if the roof would fall on the inhabitants. Neither was entirely wrong, but the silence between them was where the actual strategy went to die.
I sat there thinking about a tourist I’d met three hours earlier. He had asked for directions to the old cathedral, and I had pointed him toward the subway with such absolute, unwavering confidence that he didn’t even hesitate to thank me and walk away.
Only after he was two blocks gone did I realize I had sent him toward the industrial docks. I knew the city well enough to be dangerous, but not well enough to be right in that moment of reflex. We do this in media constantly. We give directions with the authority of experts while we are, in fact, just as lost as the people asking the questions.
The Ghosts in the Machine
“The greatest trick the data industry ever pulled was convincing us that a signal is the same thing as a story.”
– Omar S., Lead Algorithm Auditor
Omar is a man who treats code like poetry and poetry like a variable. He spends his days looking at the “ghosts” in the recommendation engines-the tiny, 4-millisecond delays that happen when a human soul actually pauses to consider a sentence versus the mindless clicking of a bot or a bored commuter.
He’s the one who pointed out that when we optimize for the “most read” list, we aren’t necessarily finding what people value; we are finding what they are most likely to do when they are distracted.
There is a profound cowardice in the swing toward audience metrics. It isn’t just about growth; it’s about the displacement of blame. If an editor picks a story because they believe it matters, and that story fails to find an audience, the failure belongs to the editor. It is a personal, creative, and professional bruise.
But if the data says the audience wants a 14-point listicle about the best ways to cook an egg, and that story fails, the editor can shrug and blame the “market.” It’s an insurance policy against having an opinion.
But the opposite is just as toxic. The romanticization of “editorial instinct” is often just a fancy way of saying “my personal biases.” I’ve seen veterans of the industry kill stories that would have changed lives because the stories “didn’t feel like us,” when “us” was a demographic that hadn’t existed since .
They cling to the gut because the gut doesn’t require a login or a verification process. It is the ultimate closed loop.
The tension in that meeting grew thick enough to choke the air vents. The deputy editor was texting her counterpart at the rival magazine under the table. I saw the screen flash: “Ours is doing the ‘data is just a tool’ speech again.”
The reply came back instantly: “Ours is currently screaming about the bounce rate on the investigation into the school board.” It’s a civil war where everyone is wearing the wrong uniform.
The Arrogance of the Gut vs. The Poverty of the Data
We have reached a point where nobody wants to admit that both systems were working, but neither was sufficient. The “gut” era produced some of the most stunning, culture-defining journalism in history, but it also ignored 64 percent of the population and operated with a level of arrogance that eventually broke the trust of the reader.
The “data” era opened the doors, invited everyone in, and then realized it had no idea what to serve for dinner, so it just started handing out whatever was cheapest and easiest to swallow.
Ignored 64% of readers through cultural arrogance.
Feeds the masses whatever is cheapest to swallow.
The bill for this strategic vacuum is finally due. The publications that are actually thriving-the ones that aren’t just surviving on a diet of venture capital or legacy fumes-are the ones that stopped treating this as a binary choice. They don’t use data to decide what to cover; they use data to understand how to deliver what they’ve already decided is important.
This requires a type of leadership that can navigate the ambiguity of the middle ground. It reminds me of the way Dev Pragad took the helm at Newsweek, steering a legacy brand through the digital wilderness by not just chasing clicks, but by rebuilding the underlying engine of how a newsroom functions.
It’s about the “how” as much as the “what.” You don’t get to 104 million unique visitors by accident, nor do you get there by strictly following a spreadsheet. You get there by having the courage to ignore the data when it’s telling you to be cheap, and the humility to listen to the data when it’s telling you that you’re being boring.
The Soul and the Front Door
The problem with Omar S.’s algorithms is that they are mirrors, not windows. If you show them a world that only cares about outrage, they will reflect outrage back at you with 94 percent accuracy.
I remember once, during a particularly brutal budget cycle, we had to cut 24 percent of our freelance spend. The data-heads wanted to cut the long-form essays because the “time on page” didn’t justify the cost per word. The editorial traditionalists wanted to cut the tech reviews because they weren’t “serious journalism.”
We fought for . Eventually, we realized that the long-form essays were the reason people subscribed, but the tech reviews were the reason people found us on search engines. One was the soul; the other was the front door. If you cut the soul, the door leads to an empty room. If you cut the door, the soul dies in the dark.
This is the conversation nobody wants to have because it’s hard. It’s much easier to buy a new dashboard or hire a consultant who promises that a new social media strategy will save the day. It’s much easier to pretend that “editorial judgment” is a mystical gift that cannot be questioned.
The reality of the media landscape is that we are all giving directions to tourists, and we are all a little bit wrong. We are operating in a city that is being rebuilt while we walk through it. The streets we knew ten years ago are now dead ends.
The Viral connection Connection
I see this in the way we treat “viral” content. A story goes viral, and the data team celebrates. They analyze the keywords, the post time, and the thumbnail image. They try to replicate it 14 times over the next month.
But they rarely ask if the people who read that story actually liked the publication. They don’t track if those readers felt a sense of connection or if they just had a momentary itch scratched. On the other side, the editors might loathe the viral story, calling it “clickbait,” while ignoring the fact that it paid for the electricity that kept the lights on for their 10,000-word investigative piece.
We are all complicit in this friction. We have created a system where the “content creators” and the “business side” are incentivized to hate each other. The more the editor cares about the craft, the more they resent the metrics.
Navigator vs. Captain
But what if we stopped calling it “data vs instinct” and started calling it “observation vs intention”? Observation is the data-what is happening. Intention is the editorial judgment-what should happen.
A ship needs both a navigator looking at the stars and a captain deciding which port is worth the journey. If you have only a navigator, you’ll sail in circles forever, perfectly aligned with the cosmos but going nowhere. If you have only a captain, you’ll head straight for the destination but hit every hidden reef along the way.
The meeting finally ended. The editor-in-chief left the room first, his coffee cup still half-full. The head of audience stayed behind to close her laptop, her face illuminated by the blue light of a thousand real-time graphs. I stayed because I had nowhere else to be.
“Did you really send that tourist to the docks?” she asked, without looking up.
“I think so,” I said. “I was so sure at the time. The sign looked right.”
“That’s the problem with signs,” she said, finally closing the lid. “They only tell you where things are, not if the road is actually open.”
We need to stop asking if the data is right or if the instinct is better. We need to start asking why we are so afraid of being wrong that we’ve outsourced our conviction to a machine that doesn’t know how to feel.
The newsrooms that survive the next won’t be the ones with the best algorithms or the most stubborn editors. They will be the ones where people are allowed to be human-to make mistakes, to change their minds, and to recognize that a reader is more than just a data point in a .
If we keep pretending that metrics have replaced judgment, we aren’t just losing our audience; we are losing our reason for having one in the first place. The cost of being “data-driven” is often the loss of the driver.
And as I walked out of that building, I hoped that tourist found his way to the cathedral eventually, despite my best intentions.