SEO for a small site

Committing to content and search as the one channel does not tell you what to write. A finished sports-guides site with no traffic is not short of pages — it is short of pages that answer something a specific person is already typing. This lesson is about how to tell those apart before you spend a quarter writing, and about the one question everybody asks first — how long until this works — which turns out to have no honest answer at all, for reasons worth understanding.

Every Google page quoted below was checked on August 18, 2026. Where Google stamps a page with its own last-updated date, that date is given with the quote, because several of the things “everyone knows” about search stopped being true on a date you can look up.

Start from what the search is for, not how many people type it

Keyword tools sell you a volume number. A volume number tells you how many people type a string. It does not tell you what any of them wanted, and at your size that second question is the only one that pays.

Google frames its own job the same way. Its public explainer of ranking opens by saying that to return relevant results, “we first need to establish what you’re looking for — the intent behind your query”. The operational version of that idea — the vocabulary Google gives the raters it pays to grade results — is in §12.7 of the Search Quality Rater Guidelines (dated September 11, 2025), which says it “can be helpful to think of queries as having one or more of the following intents”:

• Know — the searcher wants to learn something. Some of these are Know Simple: a specific fact, fully answerable in a sentence.
• Do — in Google’s words, “when the user is trying to accomplish a goal or engage in an activity.” This is the searcher who is ready to act.
• Website — “when the user is looking for a specific website or webpage.” Someone who already knows where they are going.
• Visit-in-person — the searcher wants a physical place.

The distinction that matters commercially is Know against Do. Someone typing a Know query is browsing: they want to understand the offside rule, or how a substitution rotation works. Someone typing a Do query has already decided and wants to finish: they want the printable rotation sheet, the calculator, the template. A page that serves a Do query is one step from a transaction. A page that serves a Know query is one step from a closed tab, unless you built the next step into it.

The part that lets a small site in

The same guidelines describe a ladder of readings for a single query. The dominant interpretation is “what most users mean when they type the query” — and Google adds that “not all queries have a dominant interpretation.” Below that sit common interpretations, and then minor interpretations, defined as “interpretations that fewer users have in mind,” which Google further splits into reasonable minor interpretations — ones that “may help fewer users, but they still are” — and unlikely ones.

Read that as a business fact rather than a taxonomy. Google explicitly recognises readings of a query that fewer people hold and that are still worth serving well. That is the ground a one-person site can hold, because it is precisely the ground a large publisher’s economics cannot justify producing anything for. Nobody staffs a content team to serve an interpretation held by two hundred people a month. You can serve it in an evening, and you may be the only page that does.

Google’s own starter guide gives the plain-language version: “Think about the words that a user might search for to find a piece of your content… Anticipating these differences in search behavior and writing with your readers in mind could produce positive effects on how your site performs in search results,” noting that “users who know a lot about the topic might use different keywords in their search queries than someone who is new to the topic” (SEO Starter Guide, last updated December 10, 2025). The expert and the beginner are two different intents wearing the same subject.

Google is describing a system it owns and sells against

Everything above is quoted from a vendor describing its own product, and that is worth stating plainly before you build a quarter of work on it.

Google writes this guidance about a ranking system it owns and operates. Roughly three-quarters of its parent company’s revenue comes from advertising against that system: in Alphabet’s own Q4 2025 earnings slides (presentation dated February 4, 2026), Google Search & Other, YouTube ads and Google Network together came to $82.3bn of $113.8bn in quarterly revenue — about 72%. And Google declines to specify the mechanism closely enough for anyone outside Google to check the advice: it says plainly that “Google doesn’t accept payment to crawl a site more frequently, or rank it higher” (last updated December 18, 2025), and equally plainly that it will not predict what any given change will do.

None of that makes the guidance wrong. It is the only first-party account there is, and it is markedly more honest about its own uncertainty than the agencies quoting confident timelines are. It does mean you should read it as a description of the behaviour Google would like from site owners — and treat any number Google declines to give as a number that genuinely is not knowable. Hold that last clause; it is the whole of the timeline section below.

Why a guides page can win

The instinct at this size is to assume you lose on length. You do not, and Google says so in the flattest sentence in the rater guidelines:

“Highly Meets results may be any type of content: articles, short form or long form video, images, social media posts, forum discussions, etc. There is no length requirement for Highly Meets results. Sometimes what makes a result satisfying is that it is short and easy to understand.” (Search Quality Rater Guidelines, September 11, 2025)

Effort is judged relative to the format, not on a word count. The same document: “The amount of effort expected for a short video shared on social media is less than for a full-length, professionally produced documentary… but both need sufficient effort to create satisfying content for their purpose.” The 3,000-word competitor is not beating your 600-word answer because it is longer. If it is beating you, it is beating you on something else.

Three more lines from the same file are worth having in your head:

• Small sites are judged by different expectations, not lower ones. Under “The type of website”, the guidelines say quality expectations “may differ for: Small hobbyist websites vs. large corporate websites… Websites with content created by ordinary people on a volunteer basis vs. websites with content created by professionals.”
• First-hand experience is a route to high quality you actually have. §7.3: “Experience is valuable for almost any topic… From writing symphonies to reviewing home appliances, first-hand experience can make a social media post or discussion page High quality.” The guidelines also allow that “there are many types of informal expertise that may be visible” in the content itself.
• You can be the go-to source for a narrow thing. The guidelines ask raters to “consider whether it is considered a go-to, authoritative source for the type of information it is displaying” — a question about a specific topic, not about the size of the company answering it.

Google’s present-tense name for the guidance itself is “Creating Helpful, Reliable, People-First Content” (last updated December 10, 2025), and its self-assessment framing is Who, How and Why: make authorship self-evident, disclose how the content was created — “especially regarding automation and AI use” — and be able to say the page exists “primarily to help people.” The four qualities its systems are said to look for are experience, expertise, authoritativeness and trustworthiness, with “trust is most important.”

The honest timeline, which is that there isn’t one

You will read “6 to 12 months” everywhere. It is not Google’s number. No current Google documentation states it, or any equivalent window — not the starter guide, not the core-updates page, not the traffic-drop debugging guide, not the crawling and indexing FAQ, not “How Search Works,” not “Do you need an SEO?”. It is industry folklore that has been repeated long enough to sound sourced.

There is one Google-attributed figure in circulation — “four months to a year” — and it comes from a Google Search Central video published on February 14, 2017. That video is not cited anywhere in Google’s current documentation. The old help page that hosted it now redirects to “Do you need an SEO?” (last updated June 5, 2026), which embeds no video and states no timeline at all. The only timing language left on that page is a question Google suggests you ask a prospective SEO: “What kind of results do you expect to see, and in what timeframe?” Google has moved from answering the question to telling you to ask it. The video’s transcript could not be retrieved when this lesson was checked, so treat even that figure as secondhand: its existence and date are verified, its wording is not.

Here is what Google does say, quoted rather than paraphrased.

• The range is deliberately enormous. “Every change you make will take some time to be reflected on Google’s end. Some changes might take effect in a few hours, others could take several months.” (SEO Starter Guide, last updated December 10, 2025)
• There is no promise of any effect at all. “There’s no guarantee that changes you make to your website will result in noticeable impact in search results.” (debugging search traffic drops, last updated December 10, 2025; the same sentence appears on the core updates page, same date)
• The re-check cadence is weeks, not days.“You’ll likely want to wait a few weeks to analyze your site in Search Console again,” and, in the starter guide, “in general, you likely want to wait a few weeks to assess whether your work had beneficial effects.”
• Where Google does commit to “months,” it is about content quality specifically. “It may take months before our systems determine that a site is now producing helpful content.”
• Recovery from a core update is gated on the next core update, not on a calendar. “If it’s been a few months and you still haven’t seen any effect, that could mean waiting until the next core update.”
• Google will not even predict indexing. “In general, we cannot make predictions or guarantees about when or if your URLs will be crawled or indexed” (crawling and indexing FAQ, last updated December 10, 2025), and “indexing isn’t guaranteed; not every page that Google processes will be indexed” (How Search Works, last updated December 18, 2025).

So the honest position is not a softer number. It is that the company that runs the system refuses to give one, and gives instead a range spanning three orders of magnitude with an explicit disclaimer attached. Any confident month-count you read is a vendor’s estimate, and the bias runs one way: longer quoted timelines justify longer retainers. A range that conveniently starts at six months is being quoted by people who bill monthly.

What this changes practically: pick your judge-on date by how much evidence you will have accumulated, not by a number you read. Google does say what a small site can rely on — that “sites that follow the Search Essentials are more likely to show up in Google’s search results,” likelihood rather than entitlement — and that traffic moves for reasons you do not control, including seasonality and “external influences.” Search is still the archetypal compounding channel. It is also the one where you are least able to tell, early, whether it is compounding or simply silent.

What a one-person site should not bother competing for

Four things are not winnable, and one of them is a policy violation rather than merely a bad bet.

Volume

Out-publishing a company with a content team is not a strategy you are executing badly; at scale it is a named spam violation. Google’s spam policies (last updated May 15, 2026) define scaled content abuse as “when many pages are generated for the primary purpose of manipulating search rankings and not helping users,” and give as an example “using generative AI tools or other similar tools to generate many pages without adding value for users.” Note precisely what that does and does not say: the policy targets scale without value, not the tool. Using a model to help write a page you would have written anyway is not what is being described. Generating two hundred keyword-variant pages over a weekend is.

Length

Already covered, but it belongs on this list: “there is no length requirement for Highly Meets results.” Padding a 600-word answer to 2,500 words to match a competitor is spending your evenings buying something Google says it is not selling.

The head term

The query with the dominant interpretation and the highest volume is contested by everyone with a budget. The winnable ground is the minor interpretation — the reading fewer users have in mind, which Google explicitly says may “help fewer users, but they still are.” That is not a consolation prize. It is the only segment of the market where being one person is an advantage rather than a handicap.

Technical SEO theatre

The crawl-budget rabbit hole is not yours to fall into, and Google says so in the first paragraph of the guide about it: “if your site doesn’t have a large number of pages that change rapidly, or if your pages seem to be crawled the same day that they are published, you don’t need to read this guide” (managing crawl budget, last updated July 22, 2026). It is scoped to sites of a million-plus pages, or ten thousand-plus changing daily. Site organisation only starts to matter “if you have more than a few thousand URLs” (starter guide). A sitemap “does not guarantee indexing or increase your site’s ranking” (crawling FAQ). And to publish at all, “you usually don’t need to do anything except publish your site on the web” (starter guide). Anyone selling you a technical audit for a forty-page site is selling you a document, and Google’s own answer to the ranking-guarantee pitch is that “no one can guarantee a #1 ranking on Google.”

What this does not cover

Search is patient and silent, which is exactly why it cannot be your only move while you wait to find out whether it worked. The other half of a content-and-search channel is showing up where people already are and talking about what you are building — which most solo builders do badly enough to get muted, blocked, or quietly ignored. How to post about your own work in public without reading as a spammer is what comes next.