The Complete Guide to SEO for a Small Business

Most small-business SEO fails before a word is written, on target selection. Someone picks terms by search volume, writes good content for queries dominated by national brands, and concludes after four months that SEO doesn't work for businesses their size.

It does. It just requires picking fights you can win.

The sequence

| Step | Question | Failure if skipped | |---|---|---| | 1. Choose targets | What can we realistically win? | Months writing for unwinnable terms | | 2. Read intent | What kind of page does this query want? | Right topic, wrong format, no ranking | | 3. Write the page | Does it serve the query directly? | Buried answers, weak titles | | 4. Link it | Does the site structure support it? | Orphan pages, no authority flow | | 5. Diagnose | Why isn't this working? | Rewriting pages that were fine |

Step 1 — Pick terms you can win

Start from problems, not from a tool. What clients ask you, what recurs in support, what people say in the communities where your buyers gather. Tools expand a list; they can't originate one.

Then judge winnability before volume by doing the single most useful free check available: search the term and look at page one. All major brands with recent thorough content means not yet. Thin, old results or forum threads mean genuinely winnable.

Long-tail wins earlier. Ten specific terms you rank first for beat one broad term where you sit on page four. Specific queries also convert better, because specificity means a further-along buyer.

Then group terms into clusters and write one strong page per cluster, rather than a thin page per keyword competing with your own other pages.

Score on winnability first, commercial value second, volume last — volume is the number everyone sorts by and the least useful of the three.

Full method: keyword research that doesn't waste a year.

Step 2 — Match the page type to the query

The most common reason good content ranks for nothing: it's the wrong type of page.

Search the term and read page one. Long guides mean informational. Comparison tables mean commercial. A calculator ranking means the tool is the content.

Format is part of intent. If eight of ten results are listicles, a 4,000-word essay underperforms however good it is. Serve the format, then be better within it.

Mixed results are the opportunity. When page one contains several page types, search engines are hedging — which means no single type dominates and a page serving one segment very well can take a position. That's the best situation available to a small site.

And intent shifts. If a page that used to rank has slipped, check whether page one still looks like your page before adding words.

Full method: search intent.

Step 3 — Get the eight on-page things right

Most checklists have forty items; thirty stopped mattering years ago. In order of impact:

1. Title tag — primary term near the front, written for a human, unique across the site. Highest leverage single element, and it's what appears in AI citations.
2. Answer in the first hundred words. Readers and AI summaries both take the answer from near the top.
3. Heading structure — one H1, then H2s written as the sub-questions people search. Those headings get extracted into snippets.
4. URL — short, readable, never date-stamped.
5. Internal links — up, across, and down.
6. Images — descriptive filenames, real alt text, compressed.
7. Meta description — no direct ranking effect, real click-through effect.
8. Freshness — a visible updated date, and actually updating it.

Ignore: keyword density, the meta keywords tag, exact-match repetition, and word-count targets. Padding to hit 2,000 words makes a page worse.

Full checklist: on-page SEO.

Step 4 — Link the site properly

Backlinks require someone else to decide. Internal links are the only ranking factor you fully control, and most small sites use them for navigation and nothing else.

Structure as hub and spoke: a strong page per topic, supporting pages linking up into it, the hub linking down to all of them, and siblings linking across.

Two rules that carry most of the value:

• Descriptive anchor text. "How to price your services", never "click here". The anchor tells search engines what the destination is about.
• Contextual links in the body, not only in navigation. Navigation links appear on every page and get discounted accordingly.

Then find your orphans — pages with no internal links pointing at them are effectively invisible however good they are. Publishing isn't finished until existing pages point at the new one.

Full structure: internal linking.

Step 5 — Diagnose before you rewrite

When a page isn't performing, work in order and stop at the first real cause:

1. Is it indexed? Search site:yourdomain.com/url. A surprising number of pages fail here, and nothing else matters until it's fixed.
2. Is the intent right? Wrong format is a rewrite, not an edit.
3. How old is it? Three weeks with no ranking is normal, not a diagnosis.
4. Where does it rank? Positions 11–30 are your best opportunities — already recognised, small improvements move them.
5. Is the term winnable? If page one is all major brands, change target.
6. Is it thin against what ranks? Coverage, not word count.
7. Does anything link to it?
8. Has it decayed — or is another of your own pages cannibalising it?

Most people start at step six and rewrite. That's the most expensive step and it's the correct fix perhaps a quarter of the time.

Full diagnostic: why your content isn't ranking.

What to expect

Honest timelines, because unrealistic ones cause bad decisions:

| | Realistic | |---|---| | New site, first rankings | 3–6 months on long-tail terms | | New page, established site | Weeks to a few months | | Competitive head term, new site | A year, if at all | | Improving a page ranking 11–30 | Weeks |

The compounding matters more than the speed. Pages that rank keep earning, and each one makes the next slightly easier by adding authority to the site.

What this connects to

• Content quality — ranking gets you seen; the page still has to be worth reading. That starts with genuinely answering the question rather than filling space.
• Conversion — traffic that converts nothing is a vanity metric. Match the ask to the intent: email capture on informational pages, a real recommendation on commercial ones.
• The refresh cycle — decay is real, and a page that used to rank and slipped needs a different fix from one that never ranked.

The mistakes, collected

1. Sorting targets by volume. The top terms are the ones you can't have.
2. Not checking page one before writing. Wrong format, no ranking, whatever the quality.
3. Judging a page after three weeks. Not enough time to mean anything.
4. Ignoring positions 11–30. Your best opportunities, already recognised.
5. Only navigation links. The free lever, unused.
6. Rewriting before checking indexation. The page might be invisible.
7. Publishing a page nothing links to. An orphan is invisible however good it is.

Where to start this week

Two things, neither of which involves writing anything new.

Open Search Console and find pages ranking between 11 and 30. Those are recognised and not yet winning — improving three of them beats publishing three new articles this quarter.

Check your ten most important pages for inbound internal links. Any that only have navigation links pointing at them are being starved, and fixing that takes an hour.