What is SEO and how does it work: the mechanic's view
A client once told me her site was "invisible." She wasn't wrong. She ranked on page four for her own brand name, while a competitor with half her inventory sat in the top three for every phrase that mattered. She had a website. She just didn't have SEO. And most people explaining what SEO is to her started with the definition, then stopped right where the useful part should begin.
So let's skip the textbook version. SEO — search engine optimization — is the practice of making a website easy for search engines to understand, easy to trust, and easy to recommend to the right person at the right moment. That's the whole thing in one sentence. The interesting part is what happens between "understand" and "recommend," because that's where almost everyone gets it wrong.
Key Takeaways
- SEO stands for search engine optimization — improving how your pages are found, read, and ranked by search engines.
- It works in stages: crawling, indexing, evaluating, then ranking. Break any stage and the rest collapses.
- You cannot "optimize" content a crawler never reaches or never stores. Technical work comes first, always.
- Rankings are relative. You don't win by being good; you win by being clearer and more trustworthy than the ten pages next to you.
- Expect months, not weeks. A page that starts ranking in 48 hours is an outlier, not a strategy.
What SEO actually stands for (and what it doesn't)
SEO stands for search engine optimization. That's the definition, and it takes four seconds. The confusion starts with the word "optimization" — people hear it and picture keywords sprinkled into paragraphs like salt.
It isn't that. Not anymore, and honestly not for a while in my experience.
Optimization means removing friction between three parties who all want different things: the crawler that wants to read your page efficiently, the algorithm that wants to rank the most useful result, and the human who typed six words into a search box at 11 p.m. hoping to solve a problem before they fall asleep. When those three align, you rank. When they don't, you can write the most elegant copy on the internet and watch it sit on page six forever.
How to do SEO in business without the jargon
In a business context, SEO is a distribution channel with one unusual property: it compounds. Paid ads stop the moment you stop paying. A well-built page keeps earning visits while you sleep, for years. That asymmetry is the entire business case, and it's why I tell founders to treat SEO like inventory rather than marketing spend.
Concretely, that means three recurring jobs:
- Deciding which questions your business should answer better than anyone — that's keyword and topic research
- Building pages that answer them completely, then linking those pages together so nothing is orphaned
- Earning references from other sites, because a link is still the closest thing search engines have to a vote of confidence
Miss the third one and you're shouting in an empty room. I watched a site with excellent content plateau for nine months because nobody had ever linked to it except its own author. Adding twenty-odd genuine citations from relevant sites took its organic traffic up roughly 60% in four months. Same content. Different credibility.
How search engines work, step by step
Before you optimize anything, you need the pipeline in your head. It has four stages, and they run in strict order.
Crawling: the discovery stage
Search engines send automated programs — crawlers, sometimes called spiders or bots — that follow links across the web. When a crawler lands on your site, it reads the HTML and queues every internal and external link it finds. If a page has no path leading to it, the crawler may never arrive. This is why a page buried behind a broken navigation menu effectively doesn't exist, no matter how good the writing is.
Indexing: the storage stage
Once a crawler has read a page, the engine decides whether to store it in its index — a colossal database of pages it's willing to show. Not everything gets in. Thin pages, duplicates, and pages blocked by a stray line in your robots file get dropped. I once spent two weeks tweaking headlines on pages that had never been indexed at all. Nothing I did to the copy could have worked, because the copy wasn't in the database. That was a humbling Tuesday.
Ranking: the judgment stage
When someone types a query, the engine retrieves the matching documents from its index and orders them. The ordering is the result of hundreds of signals weighed together — relevance to the query, the depth and freshness of the content, how quickly the page loads, whether it's usable on a phone, and how many credible sites point to it.
And this is the part most explanations skip: the engine also evaluates how people respond. If everyone who clicks result number one bounces back within eight seconds and picks number four instead, that pattern feeds back into future rankings. Your position isn't a fixed prize. It's a live negotiation.
What SEO looks like with real examples
Abstract advice is useless, so here are two situations I've actually worked through.
| Scenario | Problem | What fixed it | Time to see movement |
|---|---|---|---|
| Small e-commerce store | Product pages ranked nowhere; identical descriptions copied from a supplier feed | Rewrote descriptions with original detail, restructured category pages, added internal links from a popular blog post | Seven to ten weeks |
| B2B service site | Strong traffic, no inquiries — ranking for curiosity phrases rather than buying phrases | Retargeted content toward comparison and pricing questions, trimmed 30 thin pages | About one quarter |
Notice what neither fix involved: adding keywords for the sake of density. The e-commerce store didn't need more words. It needed different words — its own. The service site didn't need more traffic. It needed traffic with intent.
Which brings up the question I get asked most often.
How to do SEO for a website step-by-step
The order matters more than the individual steps. Here's what I follow, and what I'd tell you to follow even if you're doing it alone:
- Audit the technical base. Can crawlers reach every page? Is anything accidentally blocked? Is the site fast enough on a mid-range phone?
- Map the questions your audience actually asks, grouped by topic rather than by single keyword
- Build one strong page per topic instead of five weak ones competing with each other
- Optimize on-page elements: a title that matches the query, headings that follow a logical order, descriptive links
- Earn references from sites that genuinely relate to your field
- Measure what changed, then repeat — this never really ends
Step three is the one people resist. Splitting a topic across five thin pages feels productive. It usually just creates five pages that each rank twelfth.
Why nothing happens for months, and why that's normal
The single most common reason people quit SEO is timing. They publish, wait three weeks, see nothing, and conclude it doesn't work.
Here's the honest picture: a new page typically takes weeks to be crawled and indexed, then further weeks to be evaluated against competitors who already have history. I've had pages start moving in under two weeks when they targeted something genuinely underserved. I've also had pages sit flat for five months before climbing, purely because the topic was crowded with established pages. Total waste of my patience, but not of the work.
And one warning I'd give anyone: rankings you win today can soften after a major algorithm update. Search engines periodically recalibrate how they weigh signals, and sites built on thin content or manipulative link schemes tend to lose ground in those moments. Sites built on being genuinely useful tend to hold, or gain. I've gone through this cycle enough times to stop panicking about it — but I'll admit I still check the analytics more than I should on update days.
Where SEO fits against everything else
SEO is slow, cheap per visit once it's working, and fragile if you stop maintaining it. Paid search is instant, expensive forever, and vanishes the second your budget does. Social distribution is great for reach, poor for people searching with intent to solve a specific problem.
My opinion, and I'll defend it: for any business with a product or service people actively search for, SEO is the channel with the best long-term return — provided you commit for at least two quarters. Under that horizon, you're better off spending the money elsewhere.
But here's the thing that stays with me. The client who told me her site was invisible hadn't been doing anything wrong. She'd been doing things in the wrong order. She fixed the technical foundation, rewrote twelve product pages in her own voice, and earned a handful of real links from suppliers. Eleven months later she was ranking first or second for most of her category phrases, and she'd cut her ad spend by about a third.
So when someone asks what SEO is and how it works, the accurate answer is boring and true: you make your site readable, you make it trustworthy, and you keep doing that long after it stops being exciting. The engine does the rest. It just doesn't do it on your schedule.
