On-Page SEO: How to Optimise a Page Properly

Most business owners meet on-page SEO as a checklist. Put the keyword in the title. Put it in the H1. Put it in the first hundred words. Tick, tick, tick, and then nothing happens, because a checklist tells you where to put a keyword but never tells you whether the page deserved to rank in the first place.

The version worth paying for is a different exercise. You decide what one question the page answers, for whom, at what stage of their decision. Then every element on the page either supports that answer or gets cut. The keyword placement falls out of that decision almost automatically, which is why the checklist feels so easy once the thinking is done, and so pointless when it is not.

Google is unusually open about this part. Its documentation on how title links get generated spells out that it reads the title element, the visible heading, prominent styled text, and anchor text pointing at the page. When those sources disagree with each other, Google rewrites your title for you. That is on-page SEO in one paragraph. The video below covers the mechanics before we get to the decisions.

PART 1 · DIAGNOSE

What On-Page SEO Actually Controls

IN BRIEFOn-page SEO controls relevance, whether Google reads your page as the right answer to a query. It does not control eligibility, which is technical SEO, and it does not control authority, which is links. Three levers, three budgets, and confusing them is how money gets spent in the wrong place.

The clean way to hold the three apart: technical SEO decides whether your page can rank, on-page SEO decides what it ranks for, and authority decides how high. A page can be flawless technically and still rank for nothing, because nothing on it says clearly what it is about.

What sits inside the on-page lever is narrower than most proposals suggest:

• The title element. Still the single strongest on-page statement of topic, and the one Google most often overrides when it disagrees.
• Headings and content structure. The H1 and the H2s that break the page into answerable parts.
• The body copy. Whether the page covers the question fully enough that a reader stops looking.
• Internal links and their anchors. What the rest of your site says this page is about.
• Images and their alt text. Descriptive text where the page carries meaning in pictures.
• Structured data. The machine-readable summary that decides which result formats you qualify for.

PART 2 · DIAGNOSE

One Page, One Question, the Rule Behind Everything Else

IN BRIEFEvery page should answer exactly one question for one kind of reader. Pages that try to answer three end up ranking for none, because the strongest signal on the page keeps changing. Settling the question is the job that keyword research exists to finish.

Test any page you own with two questions. What one thing would a visitor be searching for when this page is the right result? And could a competitor’s page answer that same thing better in one sentence? If the first question takes you more than a few seconds, the page has a scope problem, not a keyword problem.

Scope problems show up in three familiar shapes on Malaysian SME sites:

1. The everything page. One “Services” page covering six services. It ranks for none of them, because each service dilutes the others.
2. The duplicate pair. Two pages answering the same question, often an old blog and a newer service page, quietly competing with each other for the same query.
3. The mismatched page. The title promises a comparison, the body delivers a sales pitch. Visitors leave, and Google eventually notices they leave.

The fix for the first is splitting, for the second consolidating, and for the third rewriting the body to match the promise. None of them is expensive. All of them are decisions somebody has to make before a writer opens the document.

BENCHMARK BRIEFING 1 OF 4

What Does Google Actually Document About On-Page Elements?

IN BRIEFA surprising amount of on-page advice is folklore. Google publishes its actual position on titles, snippets, images, links and structured data, and where it is silent, that silence is itself useful when judging an SEO proposal.

The table below sets each element against what Google’s own documentation says, so you can check a recommendation against the source rather than against a supplier’s confidence.

PART 3 · DESIGN

Matching the Page to What the Searcher Wants

IN BRIEFIntent decides page format before a word is written. A comparison query wants a table, a “how much” query wants numbers, a “near me” query wants a location and a phone number. Guess wrong and the page loses to a worse-written competitor that guessed right, which is why format is a design decision as much as a writing one.

The cheapest research in SEO is looking at what already ranks for your query. Google has already decided what format satisfies that search; the top ten results are that decision made visible. Use the box below to translate what you see into what to build.

BENCHMARK BRIEFING 2 OF 4

Which On-Page Fixes Are Worth the Effort?

IN BRIEFOn-page fixes vary enormously in payoff per hour. Intent alignment and title rewrites sit at the top; alt-text tidying and meta-description polish sit near the bottom. Sequencing by payoff is what separates a plan from a task list, and it is how the work gets scoped and budgeted.

Read the bars as a priority order, not a promise. The point is the ranking between fixes, which holds fairly consistently, rather than any individual figure.

PART 4 · DEPLOY

Writing the Page So Both Readers and Machines Can Use It

IN BRIEFWrite the answer first, then the reasoning. Both human skimmers and AI answer engines take the first clear statement under a heading, so burying the verdict in paragraph four costs you the citation. The same habit runs through everything in the SEO audit checklist.

Six habits do most of the work on the page itself, and none of them require a copywriter:

• Answer in the first two sentences under each heading. Then explain. Never the other way round.
• Phrase headings the way people search. “How much does it cost?” beats “Investment considerations” every time.
• Break enumerations into lists and comparisons into tables. Buried prose gets skimmed past and rarely quoted.
• Use the keyword naturally and its variants freely. Repetition beyond what reads normally is named in Google’s spam policies.
• Link out to your own related pages with descriptive anchors. “Read more” tells nobody anything.
• Write alt text that describes the image. If the picture carries meaning, say what it shows.

This matters more each year, not less. Google now publishes specific guidance for generative AI surfaces. The advice is the same on-page SEO discipline pointed at a machine reader: clear structure, unambiguous statements, and content that stands on its own when a fragment is lifted out of context.

BENCHMARK BRIEFING 3 OF 4

How Has On-Page SEO Changed Over the Years?

IN BRIEFOn-page practice has moved from placing keywords, to covering topics, to structuring answers machines can lift. Advice written for one era still circulates in the next, which is why some “SEO checklists” still sound like 2015. Our marketing insights track the drift.

The point of the timeline is to date the advice you are being given. If a proposal talks about density targets, it is describing a discipline that stopped existing some time ago.

BENCHMARK BRIEFING 4 OF 4

What Changes On-Page by Page Type?

IN BRIEFA blog, a service page and a pricing page need different on-page emphasis. Applying one template to all three is the reason many sites have technically tidy pages that convert nothing, and it is a planning failure rather than a writing one.

Use the grid to set the emphasis before briefing each page, so the writer knows which element carries the weight.

PART 5 · DRIVE

How to Tell Whether the On-Page Work Worked

IN BRIEFMeasure at the page level, not the site level. Compare the query set a page attracted before and after: winning better-matched queries is the real signal, and it usually shows before position does.

Open Search Console, filter to the single page you changed, and look at the queries rather than the average position. Three outcomes are worth naming:

1. New, better-matched queries appear. The change worked, even if position has not moved yet. Give it another month.
2. Same queries, higher position. The page was already right on scope; you improved its expression of it.
3. Nothing changes across two months. The constraint is not on the page. Look at eligibility or authority instead of editing further.

The third outcome is the valuable one, because it stops you paying for a fourth rewrite of a page that was never the problem.

EXPLORE · SEO

Where to Go From Here

Pick the page matching the decision in front of you, or start with a diagnosis and let the findings choose.

1. SEO Services Malaysia, which SEO work is worth funding, and how that call gets made.
2. Keyword research guide for Malaysian businesses, how the one question per page gets chosen.
3. Technical SEO basics every Malaysian site needs, the eligibility half that comes first.
4. SEO audit checklist: 12 fixes that move rankings, the wider audit this work sits inside.
5. IZI Digital Marketing, the consulting-first approach behind every recommendation above.

FAQ

Frequently Asked Questions

1. What is on-page SEO in simple terms?

It is everything on the page itself that tells search engines what the page is for. How much of it you need depends on your competition, in a quiet niche a clear title and honest content may be enough, while a crowded one demands full alignment of format, structure and schema.

2. How many times should a keyword appear on a page?

As often as it reads naturally, and no more. There is no target density, and deliberate repetition beyond natural use is named in Google’s spam policies. If you are counting occurrences, you are working on the wrong problem.

3. Does the meta description affect rankings?

Not directly, though it still earns its place. It shapes the snippet people read before clicking, and Google frequently writes its own from the page anyway. Write it as ad copy for the result, not as a ranking lever.

4. How long should a page be for good on-page SEO?

Long enough to answer the question completely, then stop. There is no word count Google rewards. What matters is whether a reader can leave satisfied, padding a thin page to hit a number makes it worse, not better.

5. Can on-page SEO alone get me ranking?

Sometimes, in low-competition niches or for long-tail queries. For competitive commercial terms it rarely can on its own, because relevance without authority still loses to relevance with it. Get on-page right first anyway, it is the cheapest of the three levers.