"We are posting. We've been at it for six months. Hardly anyone's reading."
We've heard this a lot over the past year or two. Posting frequency hasn't actually dropped; with AI helping write, keeping up three posts a week is manageable. But the numbers on the back end haven't moved, and inquiries haven't gone up either.
The owner's first instinct is usually "we just haven't cracked the viral formula yet." So they start studying why other people's posts take off, how to write a headline, what the cover image should look like. Keep going in that direction for another six months, and the numbers will most likely look the same.
Because the sixty posts that went out over those six months have nothing to do with each other. Sixty fresh starts from zero still add up to zero.
01Writing got cheap. Being seen got expensive.
We're fortunate to live in this era of computers, the internet, and AI. Writing an article has never been this cheap.
Our earlier pieces on GEO dealt with the mechanics: SEO isn't dead, it became the entry ticket to getting cited by AI; when a customer asks AI "who's best at this," the three names in the answer are where you win or lose; AEO, GEO, and LLM SEO are roughly the same horse in three different coats.
This piece deals with what comes after the mechanics. Once you have the ticket, what does the content you put up need to look like for anyone to want to cite it?
And AI has made that harder. When every competitor can turn out five reasonably fluent articles a day, fluency on its own stops being the bar. Search engines and AI have to pick the few worth citing out of a pile of drafts that all look alike, and the standard only gets stricter: who's specific, who has firsthand experience, who has kept talking about the same subject for three years.
When something cheap becomes plentiful, it stops being worth much. Content gets no exemption from that rule.
02Chasing viral posts is a way of dodging the question
On the surface, chasing the next viral post looks proactive. In practice, it swaps a hard question, "what do we actually want to leave in our customers' minds over these six months?", for an easy one: "how do we get traffic on this post?"
Fragmented posting costs you on three levels, and none of them show up in your back-end numbers.
The cost in recognition. Customers can't remember you. This week it's AI, next week it's office life, the week after that you're reposting a piece of industry news. Each post is fine on its own; read together, they don't let a reader piece together what you actually do or what you're good at.
The cost in efficiency. You start over every day. Without a through line, topic, material, and tone all have to be decided from scratch every time. Six months in, the only thing that's built up is fatigue.
The cost in assets. Nothing is left behind. Traffic from a viral post is a one-off: it looks good this month, drops to zero next month, and doesn't make the next post any easier to write.
Fragmented content is pitching a tent; systematic content is building a house. A tent goes up fast, but one rainstorm and you're packing it away. A house goes up slowly, but in the second year you're building on top of what you finished in the first.
03Sounding like AI isn't a writing problem. It's having no point of view.
"How do I make AI-written articles not sound like AI?" is probably the most-searched content question right now.
The symptoms are easy to spot, and come in roughly three kinds. Empty: every sentence is correct, but there's no scene, no action, no result. Stuffed: all summary and grand framing, and you can't remember a single line afterward. Smooth: too neat, like a model answer, not like someone talking.
There are plenty of fixes online, and most of them genuinely work. Cut the vague, lofty lines. Swap press-conference words like "empower," "boost," and "build out" for plain language. Add concrete scenes and results. Flatten the template sentence patterns. And whenever you see "industry insiders believe," ask who exactly said it.
The problem is that these five fixes repair the wording, not where the content comes from.
Sounding like AI has only one root cause: there's nothing in the text that only you would know.
What a model can write is the average of everything it has read. Ask it to write "how small businesses can go digital," and it compresses ten thousand articles on the same topic into one reasonably fluent average draft. You can't find anything wrong with that draft. There's just nobody in it. And readers can tell at a glance, because they've read a few of those ten thousand too.
So why is a sentence like "industry insiders generally believe" especially fatal? Its problem isn't precision. It's actually a confession: I haven't looked into this, and I haven't done it either; I'm just talking as if I know.
Flip it around, and the material only you have is something AI can't write a single word of. What a client asked last week, which approach you tried that failed, which line on your quote gets cut most often. Those are the raw ingredients that make an article sound human, and they have nothing to do with writing skill.
Let AI move fast. Let people hold the anchor. AI produces the rough draft; people do the editing and the judging. Once that division of labor is flipped, what you get is the pile of articles online right now that leave nothing behind once you've read them.
04What systematic actually looks like
Systematic sounds like a production only big companies can afford. In practice, it asks for just four things.
The first is a consistent style, so readers remember you. The same tone, the same stance, the same visual look.
The second is steady output, building up on a rhythm. The frequency doesn't need to be high, but it needs to be predictable.
The third is capability that's captured, not tied to any one person. How topics get found, how drafts get edited, what can and can't be written: all of it is written down as a process, so the result doesn't drift when someone else does the work.
The fourth is assets that build up, so every piece makes the next one easier to write. Older pieces can be picked up by new ones, cited, and broken into other formats to be used again.
These four things sound abstract described this way, so it's faster to look at an example. The piece you're reading is one.
It isn't a standalone article. It belongs to a through line: first whether to do it, then what to feed it, then who does it, and finally how to get seen. Every piece points back to an earlier one (this one picks up from four GEO pieces right at the start), and each leaves room for a later one. Every piece maps to a service we actually provide, so once it's written, we know where to send the reader next.
That's why it costs this team far less to write the nineteenth piece than it did to write the first. The nineteenth is written standing on the eighteen before it. That's what assets building up actually looks like.
05Where to start
If what you have right now is six months of posts that don't connect, don't rush to write another one. Three things come before the writing.
01 Set the through line. Pick three to five topics where you genuinely have something to say, and that relate to how you get paid. Three to five is plenty; any more and you can't hold them.
02 Take stock of your firsthand material. List what you have on hand: the questions customers ask most often, what you've learned from past projects, the disputes that keep coming up during quoting and delivery. These are raw ingredients nobody else can copy.
03 Set the order. Make every piece connect to the one before and leave room for the one after. Write the foundation topics first, then the advanced ones.
Once those three are done, most people find their content was never in short supply. What was missing was someone to put it in a line.
How to arrange that line, which topics are worth writing, which topics people are actually asking about in AI search: that's exactly what our SEO and content service does. It isn't about posting a few more pieces for you. It's about building the material you already have into a house that stands. If you'd like to see where your visibility sits right now, you can start with an AI search visibility audit, get a sense of direction, and then decide whether to break ground.
Those six months weren't wasted
Back to that line: "We've been posting for six months, and hardly anyone's reading."
Those six months weren't wasted. At the very least, those posts prove you can get things out the door. The only difference is that they're sixty tents, each pitched and packed away on its own, while not a single shovelful of earth has been turned for the foundation.
A viral post is luck, and you can't put luck on a calendar. A system you can.
Rooted in craft. Built for the new wild.