What SEO Actually Is: A Mental Model for Getting Found in Search

Imagine a library the size of the planet, with no catalogue, no librarian at the desk, and new books arriving by the millions every single day — most of them unannounced. Somebody still has to wander the shelves, open each new book, work out what it is about, file a one-line summary, and then, the instant a visitor asks a question, pull the three or four books most likely to answer it. That somebody is a search engine. And SEO — Search Engine Optimization — is simply the craft of making your book easy to find, easy to understand, and obviously the best answer to the question being asked. Before any tactic makes sense, you need the mental model of how that library actually runs. That is what this article builds.

Search Works in Three Stages

It is tempting to picture Google as a single magic box: you type a question, answers appear. But underneath, the work splits into three separate stages, and — this is the part almost everyone misses — a page can succeed at one stage and fail the next. Google states this plainly in its own documentation: not all pages make it through each stage.

1. Crawling — discovery

First, Google has to find out your page exists. There is no master list of every page on the web, so Google runs automated programs called crawlers (the main one is nicknamed Googlebot) that follow links from pages they already know to pages they don't. When a crawler reaches your page, it downloads the text, images, and other files — much like a reader pulling a book off the shelf to skim it.

2. Indexing — understanding

Downloading a page is not the same as understanding it. In the indexing stage Google analyzes what it fetched: what the page is about, what its title says, whether it duplicates a page already stored. If the page is judged useful and unique enough, its summary is filed away in the index — a colossal database of everything Google has decided is worth remembering. A page that is crawled but judged thin or duplicate can be quietly left out. Crawled does not guarantee indexed.

3. Serving — ranking the answers

Finally, when someone types a query, Google searches its index for pages that match and returns them in an order — the most relevant and trustworthy first. This ordering is what people usually mean by "ranking." Being in the index is the price of entry; ranking well is the competition that happens once you're inside.

How does it decide the order? Not by any single rule, but by weighing a great many signals at once — how well the page's content matches the query, how usable and trustworthy the page seems, and the context of the search itself. That last part explains something you have probably noticed: the same query can return different results for different people. Where you are, what language you search in, and what device you hold all feed into the order. There is rarely one fixed "position" for a page — only its position for a particular search, by a particular person, at a particular moment.

Google Is Not the Only One — but the Pattern Is Universal

So far this has all said "Google," and for a reason: it handles the large majority of searches across most of the world, and it publishes unusually detailed public documentation about how it works — which makes it the best teacher. But it is not the only search engine, and it helps to see where it sits among the others.

Microsoft's Bing runs its own crawler (Bingbot) and its own index. Yandex dominates in Russia and Baidu in China, each with its own crawler and index tuned to its language and market. Privacy-focused engines like DuckDuckGo take a different route: rather than crawling the whole web themselves, they draw heavily on another engine's index — DuckDuckGo's results come largely from Bing — and add their own layer on top. Others, such as Brave Search and Ecosia, sit at various points along that spectrum between running everything themselves and building on someone else's index.

Here is the reassuring part. Every one of them works in the same three stages you just learned: something has to discover a page, understand and store it, and then order it against rivals when a query arrives. The vocabulary barely changes from engine to engine. This is why SEO is rarely about chasing one company's quirks — make a page genuinely discoverable, understandable, and useful, and you help every engine at once. The examples that follow use Google because it is the one most people are trying to reach and the one that explains itself most openly, but the mental model is the general one, not a Google-only trick.

The Funnel: Why "Crawled" ≠ "Indexed" ≠ "Ranked"

These three stages form a funnel. Every page starts in the vast pool of "things on the web," and at each stage some pages drop out. Understanding this funnel is the single most useful idea in all of SEO, because it tells you where a problem lives when your page isn't showing up.

Read the funnel from left to right and a powerful diagnostic falls out of it. If your page brings in no visitors from Google, the useful question is never a vague "why don't I rank?" It is "which stage is my page failing?"

Most beginners assume every problem is a ranking problem and start tweaking wording. But a huge share of "invisible page" cases are actually crawling or indexing failures — the page never made it into the library at all, so no amount of polishing the prose would help. Knowing the funnel saves you from fixing the wrong stage.

You Can Talk to the Crawler

A search engine feels like a force of nature, but you are not a passive bystander. Your site sends signals at every stage, and a large part of SEO is simply sending clearer ones. One of the first things a crawler checks is a small text file at the root of your site called robots.txt, which politely tells crawlers where they may and may not go:

That last line points to a sitemap — a list of the pages you'd like search engines to know about. The deeper point is conceptual: crawling, indexing, and ranking are not things that merely happen to your site. They are a conversation, and you hold one side of it. Most of SEO is really just a way to make your half of that conversation clearer at one specific stage.

A Living Example: This Very Site

Abstractions get slippery, so let's ground the funnel in something concrete — the site you are reading right now. When you publish an article here, say From Rows to Insights: Aggregate Functions, it travels the exact pipeline in the diagram above:

1. Crawling — Googlebot reaches the article by following a link to it, perhaps from the homepage or from another SQL article that links to it. No link, no discovery.
2. Indexing — Google reads the piece, sees it is a substantial, original explanation of SQL aggregation, and files it in the index under what it's about.
3. Serving — when someone searches for how GROUP BY works, Google weighs this article against every other indexed page on the topic and decides where it lands.

Notice one detail about that link. At the time this is published, the article lives at the address /article/25. That number is perfectly functional for a human clicking a link — but it tells a search engine nothing about the topic. A crawler reading /article/25 learns about as much as you would from a book whose spine reads only "Book 25." A descriptive address such as /sql/aggregate-functions would let the address itself hint at what the page is about.

This is worth being candid about, because this site is itself a learning project, written while its author works through SEO in the open. Those numeric addresses are a real, honest imperfection — and improving them is one of the changes already on the list. So if you are reading this some time after it was published and the link above no longer lands on /article/25, that is not a broken example: it is the very lesson of this article being applied to the site you are reading.

So What Is SEO, Really?

With the pipeline in view, we can give SEO a definition that isn't hand-wavy. SEO is the practice of helping each stage do its job:

• Help Google crawl — make sure pages are discoverable and not accidentally blocked.
• Help Google index — make each page clearly about something, and worth storing.
• Help Google rank — be the most genuinely useful answer to a real question people ask.

Notice what's not on that list: tricks, keyword-stuffing, gaming a loophole. Modern search is very good at spotting manipulation, and the entire thrust of Google's guidance is that the durable strategy is to actually be the best answer. That is good news for a site like this one, whose whole premise is explaining things clearly. Genuinely useful writing and good SEO are, more and more, the same activity viewed from two angles.

The Map You Now Have

Strip away the jargon and a search engine does something almost humble: it finds pages, makes sense of them, stores them, and puts them in order when someone asks a question. Three stages — discovery, understanding, ranking — arranged as a funnel that pages must pass through in turn, where clearing one stage never guarantees the next.

That single picture is enough to change how you read every SEO problem you will ever meet. When a page brings in no traffic, you no longer face a vague mystery; you have a precise question — which stage is failing: crawling, understanding, or ranking? — and each answer points to a different, concrete fix. Almost everything else in SEO is detail hung on this one frame.