Crawl, index, rank — what actually happens
Three separate machines, and people collapse them into one, which is why so much SEO advice sounds like superstition.
Crawling. Googlebot fetches web pages, follows the links on them, and fetches those too. It finds your site through links from other sites, your XML sitemap, and links between your own pages. It doesn't discover anything that nothing points to.
Indexing. Fetching isn't storing. Google renders the page, works out what it's about, and decides whether it's worth keeping. Plenty of fetched pages are never indexed — duplicates, near-empty pages, pages that say what a hundred others said better. An unindexed page cannot appear for any query, full stop.
Ranking. When someone searches, Google doesn't go and read the web. It queries its stored index and orders the results in a fraction of a second. This is why change is slow: your edit has to be crawled, re-indexed and re-evaluated before anyone sees it.
So SEO is three jobs, not one: get fetched, get stored, get chosen. Most sites that "don't rank" are actually failing at step two and nobody checked. Search Console's Pages report tells you in ninety seconds.
The three questions a search engine asks about your page
Tactic lists are how SEO gets taught and why it gets misunderstood. Here's the frame that actually holds: every ranking decision reduces to three questions asked in order, and every real SEO task is an answer to one of them. Fail question one and questions two and three never get asked.
1. Can I fetch this page?
The plumbing question. It sounds trivial until you learn how many sites fail it silently.
robots.txtblocking a directory somebody forgot about, often the one holding your CSS or your product pages.- A
noindextag left behind after a staging site went live. This one is common enough to be a genre. - Content that only appears after JavaScript runs. Google renders JavaScript, but rendering is queued and imperfect — server-rendered content is fetched more reliably.
- Redirect chains, 404s and pages no internal link points to. If nothing links to it, nothing finds it.
- A server that takes six seconds to respond. Crawlers have budgets and they spend them elsewhere.
2. Do I understand what this page is for?
Understanding means matching your page to a query someone actually types. Not a topic — a query, phrased the way a human phrases it.
That's what titles, headings, body copy, internal links and structured data are all doing. Not tricking anything. Removing ambiguity.
- One clear primary query per page. Two pages competing for the same one is keyword cannibalisation, and it costs you both.
- Match the intent, not just the words. Somebody searching "seo cost" wants numbers. A page of philosophy about value will lose to a page with a price table.
- Internal links with descriptive anchor text tell Google what a page is about and how important you consider it. This is the most underrated free signal on any site.
- Schema markup states facts explicitly — this is a product, this is its price, this is an FAQ — rather than making the machine infer them.
3. Should I show this instead of the alternatives?
Now it's a competition. Ten pages can all be fetchable and relevant. Only one is first.
Links from other sites remain the clearest external signal, because they're the one thing you can't award yourself. A link from a site with its own audience is an editorial vote; a link from a directory that lists anyone is not, and Google has spent two decades learning the difference.
The other differentiator is first-hand substance. Real prices, real screenshots, real numbers from work you actually did — the things a generic writer can't produce because they've never done the thing. That's what E-E-A-T is pointing at, clumsily.
Organic results versus paid results
Both appear on the same page. They're bought completely differently, and mixing them up is the source of a great deal of confused spending.
- You cannot pay Google for an organic position. Not through Ads spend, not through Workspace, not through anyone claiming an "inside contact". Anyone selling that is selling nothing.
- Ads and SEO are complements, not rivals — see PPC vs SEO. Paid tells you which keywords convert in three weeks. SEO then wins those keywords without the per-click bill.
| Dimension | Paid (ads) | Organic (SEO) |
|---|---|---|
| How you get there | Bid in an auction | Earn it — no payment channel exists |
| Time to appear | Same day | Weeks to months |
| Cost per click | Every single click, forever | Zero, once you're there |
| What happens if you stop | Gone that afternoon | Decays over 6–18 months |
| Labelled? | Yes — "Sponsored" | No |
| Controllable? | Almost entirely | Influenced, never controlled |
What SEO is not
The misconceptions are durable, expensive, and worth naming individually.
- Not keyword stuffing. Repeating a phrase forty times stopped working around 2005 and now actively hurts. Write for the person; the phrasing follows.
- Not a one-time fix. Competitors publish, Google updates, content ages. A site left alone for two years loses ground it once held — see what happens when you stop doing SEO.
- Not tricking Google. The tricks that work are temporary and the ones that stop working take your site with them. There's a bar; you either clear it or you don't.
- Not the same as content marketing. Content is one of four disciplines, alongside technical, links and reporting. Buying only content and calling it SEO is why so many blogs get published and so little changes.
- Not guaranteed rankings. Nobody controls the index, and Google's own guidelines warn against agencies promising positions. A guarantee against your own baseline is honest. A guaranteed #1 is not.
- Not free. It's unpaid clicks bought with paid work — the bill goes to salaries instead of an ad auction.
Where AI answers fit in
Google's AI Overviews, ChatGPT's browsing, Perplexity — all of them changed the shape of the results page and none of them replaced the pipeline underneath. They still need content that has been crawled, stored and retrieved. The three questions are unchanged; the competition is simply harsher, because an AI answer summarises the field and cites two or three sources instead of showing ten blue links.
Practically: pages that already rank are overwhelmingly the pages that get cited, so classic SEO remains the price of entry. Beyond that, write so a machine can lift a clean answer — say it in the first two sentences, put real numbers in it, don't bury the point. That's also how humans prefer to read.
The uncomfortable part: some queries now get answered on the results page and never produce a click, and low-commercial-intent informational queries lose the most. Queries where someone is choosing a supplier still send clicks, because nobody hires a vendor a chatbot mentioned without looking first. Weight your content plan accordingly — see what AI Overviews did to organic traffic.
What it takes, honestly
SEO splits four ways: technical fixes, content, links, and measurement. A programme missing any one of the four will stall — and most stall on the one nobody enjoys, which is technical.
On timing: three to six months to real movement on a competitive keyword set, six to twelve to compounding returns. Leading indicators — impressions and average position for your target queries — should move inside four to eight weeks. Flat impressions at month three is a genuine warning sign, not a patience problem.
On money: the Indian market runs roughly ₹25,000–₹3,00,000 a month depending on site size and competition. We run SEO from ₹75,000/mo, smaller sites from ₹40,000. We never promise a position; we freeze your trailing-90-day qualified organic leads on day one and if we haven't beaten that in 90 days we keep working free until we do.
And the honest note: a small local business can do most of this itself. Fill in the Google Business Profile, fix site speed, write plainly about what you sell, ask happy customers for reviews. Agencies earn their fee in competitive markets — not where the whole job fits on one page.