SEO writing: how to write content that ranks and reads well.

The myth goes like this. There is "writing for humans," warm and clever and a little indulgent, and there is "writing for SEO," clinical and keyword-soaked and faintly embarrassing. You can do one or the other. The good news is that this framing is about ten years out of date, and Google has spent the decade dismantling it. Modern search engines are not counting keywords like a cash register. They are trying, imperfectly but relentlessly, to figure out which page best helps the actual person who typed the actual question.

That changes the whole assignment. Google's own guidance is unusually blunt about it. In its documentation on creating helpful, reliable, people-first content, it tells you to make content for people, not to manipulate rankings, and it offers a gut-check question that belongs above every writer's desk: is this primarily made to attract visits from search engines, rather than to help humans? If the honest answer is yes, you are writing the second kind of bad page. The entire skill we are about to walk through is making that answer no while still ranking. They stopped being in tension a long time ago.

Most SEO writing advice opens with keyword research, and that is exactly backwards as a starting move. The keyword is a clue. What it points to, the thing that actually matters, is intent: the specific job in the searcher's head when they typed those words. Two people can search the same phrase wanting completely different things, and the page that wins is the one that guesses right.

Take "running shoes for flat feet." One searcher wants a buyer's guide comparing five models. Another wants to know whether flat feet even need special shoes. A third wants to buy, right now. The phrase is identical; the satisfying answer is not. This is why Ahrefs' guide to SEO writing puts matching search intent ahead of nearly everything else, and why it recommends a step most writers skip.

• Read the page one results before you write a word. They are a free answer key. Google has already tested thousands of pages against this exact query and is showing you the survivors.
• Notice the format that's winning. All listicles? A how-to? Product pages? If the top ten are comparison articles and you publish a sales page, you have answered a question nobody asked.
• Notice the depth. If every ranking page covers arch support, pronation, and durability, those are not optional subtopics. They are the table stakes the searcher expects.

Get intent right and the rest of the work has a target. Get it wrong and your prose can be Pulitzer-grade and still lose to a mediocre page that answered the question you ignored.

Here is a humbling fact that should reshape how you write every page. People do not read online. They scan. This is not a hot take; it is one of the most replicated findings in web usability. The Nielsen Norman Group's research on how users read on the web found that the large majority of users scan any new page rather than reading it word by word. Your carefully built argument is being skimmed by someone with three other tabs open.

That is not a tragedy. It is a spec. Write for the scanner and you serve the careful reader too, but the reverse is not true. The moves are simple, and you know most of them.

• Front-load the answer. Say the most useful thing first, then explain. This is the inverted pyramid, borrowed from journalism, and NN/g argues it is perfectly suited to the web precisely because readers bail early. Put the payoff up top, not in a final paragraph nobody reaches.
• Use real headings. Descriptive subheads let a scanner navigate and tell search engines how the page is organized. "How much it costs" beats "A few thoughts on pricing" every time.
• Keep paragraphs short. One idea each. A wall of text is a closed door. Two to four sentences is plenty.
• Use lists where lists belong. When you are enumerating things, enumerate them. Do not bury five parallel points in a paragraph and make the reader sort them.

None of this dumbs down your writing. It respects the reader's time, the most flattering thing a writer can do.

Now the keyword part, far less mysterious than the discourse suggests. You are not seasoning the page with a target phrase. You are labeling it so the engine and the reader both know what it is about. That means the keyword in a handful of load-bearing spots and natural writing everywhere else.

Those spots are the title, the H1, the first hundred words, a subheading or two where it genuinely fits, the URL slug, and the body wherever the topic naturally calls for it. Backlinko's SEO writing guide makes a point worth tattooing on your knuckles: the location of your keyword matters more than how many times you use it. Hit the structural spots once, naturally, and you are done. There is no ideal density to chase, and chasing one is how good writers turn into stuffers.

The more sophisticated half is entities, a fancy word for the related concepts a real expert would mention. Writing about flat feet and running, you will naturally reference arch support, overpronation, stability shoes, and orthotics, not because a tool told you to, but because you cannot competently cover the topic without them. Search engines read those terms as evidence you know the subject, and you get them for free the moment you write about the thing properly instead of gaming the system. Cover the topic the way a knowledgeable friend would, and the entities show up on their own.

For years the advice was to demonstrate expertise. Google quietly added a letter to that formula, and it changed who wins. The framework is now E-E-A-T: experience, expertise, authoritativeness, and trustworthiness. That first E, experience, is the one most content ignores, and it is increasingly the differentiator. The question is no longer just whether a writer knows the subject in theory. It is whether they have actually done the thing.

Google's helpful content guidance leans hard on this, asking whether content shows first-hand expertise and a depth of knowledge that comes from having actually used a product, visited a place, or lived the experience. It is also candid that trust sits at the center of E-E-A-T, with the other qualities feeding into it. Great news for anyone with real expertise, terrible news for anyone summarizing the same ten articles everyone else summarized.

• Write what you have actually seen. The screw-up that cost a client real money, the result that surprised you, the workaround you only know because you hit the wall. Specifics that cannot be googled are the fingerprint of real experience.
• Show your work. Real numbers, real screenshots, a process you genuinely ran. Generalities are cheap. Particulars are expensive, which is exactly why they signal value.
• Make authorship clear. A real byline, a real bio, a real human standing behind the claims. Anonymous advice is harder to trust, and trust is the whole game.

You cannot fake this convincingly anymore, and that is the point. Content that reads like a person who has been there now ranks like it too, because being there is part of what is measured.

You can write the best page on the internet and still lose if your title is boring, because a result nobody clicks quietly slides down the page. The title link and meta description are the storefront, where a searcher mid-scan decides whether you are worth the tap. Resist the urge to stuff here, where the temptation peaks. Google's documentation on writing good title links tells you to keep titles descriptive and concise and warns against keyword stuffing, noting there is no reason to repeat the same words over and over. Their example of what not to do, one phrase listed four times, helps no one and reeks of trying too hard.

So write the title for the human deciding whether to click, get your keyword in once and near the front, and make a promise the page keeps. Then write a meta description that reads like ad copy, not a robotic summary. It is not a direct ranking factor, but it is the line that wins or loses the click, and clicks are the whole reason you wrote any of this.

Once they land, send them somewhere. Internal linking is the most underused move in SEO writing and one of the easiest: link your own pages to each other on purpose. It helps readers go deeper and helps search engines understand how your pages relate and which ones matter. The rules are simple. Link to genuinely related pages, the ones a reader finishing this article would actually want next. Use descriptive anchor text that says where the link goes, not "click here" or a naked URL. And do it as you write, when the connection is fresh. The goal is a site that feels like a connected library, not a pile of orphaned PDFs.

Let us address the obvious tool in the room. Yes, you can generate a competent-looking draft in nine seconds, and no, you should not ship it as written. Google does not penalize content for being AI-assisted; it penalizes content for being unhelpful, derivative, and made to game search rather than serve a person. The trouble is that raw AI output is fluent and confidently empty, the textbook case of content that reads fine and says nothing. It is the new on-ramp to that second kind of bad page, the one that ranks for a while and then quietly stops.

So treat the draft as a fast intern, not a finished writer. The value you add is what a model cannot fake: your real experience, your specific numbers, your contrarian take, the example only you have. Cut the throat-clearing and the four-adjective sentences. Verify every claim, because models invent statistics with a straight face and your name is on it. Run the read-aloud test, the single best editing habit there is, and rewrite anything that makes you wince. Use the machine to beat the blank page, then do the human work that makes the page worth ranking. That work is the moat, because it is the part nobody else will do.

Here is the honest version the listicles leave out. Everything above is real and it works, but none of it is a cheat code, and anyone promising you rankings on a schedule is selling something. Search is a moving target, your competitors are not standing still, and the single biggest factor is still whether your page is genuinely the most useful answer to the question someone asked. The techniques are how a great page gets found. They are not a substitute for being a great page.

The encouraging part is that the bar is mostly laziness, and laziness is beatable. Most content is skimmed off the same sources, optimized for a robot, and stripped of anything a real person would know. Show up with genuine experience, structure it for a human in a hurry, label it clearly, and edit like you care, and you are already ahead of most of what gets published. If you would rather hand the grind to someone who does this for a living, that work is what we do at Mining Wells under our SEO and content services. But you do not need us to start. You need a real point of view, a structure built for a scanner, and the discipline to read it out loud before you hit publish.