Module 4: On-Page SEO
On-page SEO is everything you control on the page itself: the title a searcher sees, the words and structure on the page, and the links connecting your pages to each other. It's the fastest layer to change and the one clients can watch you improve in real time. This module covers titles and headings that earn clicks, page structure that makes importance obvious, and the refresh work that quietly outperforms most brand-new content.
Titles, meta descriptions, and headings that earn the click
Your page has two jobs in the search results: rank, and then get clicked. The title tag carries both. It's the heaviest on-page ranking signal and it's the headline a searcher reads, so write it for both audiences. Put the page's main keyword near the front, then give a human a reason to pick you. "Emergency Plumber in Austin — Same-Day Service, No Call-Out Fee" will out-click "Home | Smith Plumbing LLC" every day of the week, because the first one answers the search and the second one just names a business.
The meta description doesn't move rankings directly, but it decides whether a ranking turns into a visit. You get roughly 150–160 characters before truncation. Spend them on the specific benefit, one differentiator, and a soft push: what the reader gets, why from you, and what happens next. Google rewrites a large share of descriptions using text pulled from the page itself, so don't treat that as an excuse to skip writing them. Write a good one, and make sure your opening paragraph is strong enough to serve as the backup when Google goes shopping on your page.
Headings do structural work. Use one H1 that states the page's topic plainly — it should match what the searcher came for. Then treat your H2s as a table of contents: each one a real subtopic a reader might scan for, some carrying natural variations of your keyword where they genuinely fit. A reader skimming only your headings should understand the whole page.
Here's the working habit that ties it together. In Search Console, pull the queries for a page and sort by impressions. Find the query with high impressions, a middling position, and a click-through rate well below what that position should earn. That gap is usually a title problem, not a content problem, and it's the cheapest fix in this whole course. Avoid the traps while you're at it: the same title pasted across twelve pages, keyword lists stuffed where a sentence should be, and clickbait the page can't back up. A click that bounces in five seconds is worse than no click at all — you spent the visit and taught Google the page disappoints.
A client's service page ranks around #6 for "roof repair dallas" with 4,000 impressions a month and a 0.8% click-through rate. Its title tag reads "Home — Johnson & Sons LLC." What's the highest-value first move?
Page structure and internal linking: make importance obvious
Search engines can't read your mind about which pages matter. They infer it from structure — and internal links are how you vote. Every link from one of your pages to another passes a signal: this page is worth visiting, and this anchor text describes what's there. A page linked from your homepage, your navigation, and five related articles is clearly important. A page linked from nowhere is a rumor.
Start with the page itself. Answer the searcher's main question early — the first screen should confirm they're in the right place and start delivering. Then organize the rest of the page around the sub-questions a real searcher would ask next, in roughly the order they'd ask them. Pricing factors before pricing tables, symptoms before solutions, basics before edge cases. Depth is good; wandering is not. Every section should earn its place by serving the intent you mapped back in Module 3.
Then wire the pages together deliberately. Your highest-traffic pages are distribution channels: link from them to the service and product pages that make money, using anchor text that describes the destination in plain words. When you publish something new, go back and add three to five contextual links to it from your most relevant older pages — this one habit prevents most orphan pages. And audit for orphans regularly: any page reachable only through the sitemap has effectively been abandoned, and search engines treat it that way.
Two failure modes to avoid. First, linking everything to everything: fifty links in a sidebar widget pointing at every page on the site dilutes the signal to noise, because a vote for everything is a vote for nothing. Second, manipulative anchor text — a wall of links all reading "best cheap roof repair dallas" looks engineered, because it is. Natural, descriptive anchors in sentences you'd write anyway are both safer and more useful to readers. The test for any internal link is simple: would it help a human who clicked it? If yes, it almost certainly helps the ranking too.
A client's best-converting service page gets almost no organic traffic. Your crawl shows it has zero internal links — it's only reachable through the XML sitemap. What's the most likely issue and the right fix?
Refreshing old content: the highest-return task in SEO
New content starts from zero: no history, no links, no track record. Old content already has all three — which is why improving what's already ranking on page two routinely beats publishing something new. A page sitting at position 9 with a year of history behind it is one good refresh away from position 4. A brand-new page is months away from either. When a client asks where their retainer goes furthest, this is usually the honest answer.
Build the refresh list from Search Console. Look for pages with steady impressions sitting in positions 5–20 — close enough that improvement matters. For each one, do an intent check before you touch a word: search the query yourself and read what's ranking now. Intent drifts. A page written for "learners comparing options" may be losing to product pages because searchers now want to buy. If the top results changed shape, your page has to change shape too, not just get longer.
The refresh itself: update anything stale (prices, dates, dead tools, superseded advice), add the subtopics the top-ranking pages cover that yours skips, strengthen the opening so it answers faster, and replace weak examples with better ones. Then look sideways for consolidation candidates: two of the client's pages chasing the same query, both stuck mid-page-two, are usually one strong page wearing a split signal. Merge them into the stronger URL, keep the best material from each, and point a 301 redirect from the weaker page to the survivor.
Finally, prune — honestly, and carefully. Most sites carry zombie pages: no traffic in a year, no backlinks, no business purpose. They bloat the crawl and dilute the site's overall quality signals. Before removing anything, check it for inbound links and any conversions; a quiet page with real links gets redirected to the nearest relevant page, not deleted. What's truly worthless can go. Put the whole cycle on a quarterly rhythm — pull the list, refresh the top candidates, consolidate the collisions, prune the dead weight — and a client's back catalog becomes a compounding asset instead of a junk drawer.
Two posts on a client's site both target "how to winterize a pool." They rank #14 and #17, and neither has moved in months. What's the best move?
What you can now do
- Write a title tag that carries the main keyword near the front and still gives a searcher a concrete reason to pick your page.
- Use page structure and internal links to vote for your most important pages, instead of leaving orphan pages to be read as unimportant.
- Spot a page-two ranking worth refreshing and improve the page with history behind it before commissioning anything brand-new.