There are two kinds of SEO problems. One kind lives out in the world: who links to you, what Google has indexed, how you rank for a given query. The other kind lives right there in your page's HTML, and you control every character of it. This guide is about the second kind.
On-page elements are the most fixable problems in all of SEO. A bad title tag isn't a months-long content campaign — it's a one-line edit you can ship before lunch. The catch is that almost nobody audits them deliberately. They get written once when the page is built, copied from a template, or left to whatever the CMS generated by default, and then never looked at again. That's how a 40-page site ends up with 40 identical title tags and a homepage H1 that says "Home."
This is the on-page companion to the broader DIY SEO Audit Checklist, which covers crawling, indexing, and the Search Console workflow. Here we're staying on the page itself: titles, metas, headings, alt text, canonicals, URLs, and the duplication that quietly happens across all of them.
By the end you'll know:
- How to write a title tag that actually earns the click — and what length gets it truncated
- The real rules for meta descriptions (including the one Google ignores)
- How to fix broken heading hierarchy and the "multiple H1" myth
- Where thin content and duplicate titles hide, and how to spot them across your whole site
- Which on-page issues you can fix today and which need a tool to find at scale
Part 1 — Title Tags: The Single Highest-Leverage Element
Your title tag is the blue link in Google's results and the text in the browser tab. It's the most important on-page SEO element you have, because it does double duty: it tells Google what the page is about, and it's the headline a searcher decides to click (or not).
1. Check that every page has a unique, intentional title
Open a page, view source (Ctrl+U), and find the <title> tag. Or just look at the browser tab.
What "bad" looks like: The title is your brand name on every page ("Acme Plumbing"). Or it's the CMS default ("Home | WordPress Site"). Or it's empty and the tab just shows the URL. Any of these means Google is guessing what your page is about.
What to do: Every page gets a title that describes that page. A good pattern for most pages is Primary Topic — Secondary Detail | Brand. A service page might be Emergency Drain Cleaning in Austin | Acme Plumbing. Lead with the words a searcher would actually type.
2. Keep titles to roughly 50–60 characters
Google truncates titles in search results at around 600 pixels — which works out to about 50–60 characters for most fonts. Past that, your title gets cut off with an ellipsis, and whatever you put at the end disappears.
What "bad" looks like: A 90-character title where the useful keyword is sitting at the very end, invisible in the actual search result. Or front-loaded brand name eating the first 20 characters before you say anything specific.
What to do: Front-load the important words. Put the page's real subject in the first 50 characters and the brand name last, where truncation does the least damage. If you're over 60 characters, tighten it — you almost never need the filler.
3. Stop stuffing keywords
What "bad" looks like: Plumber Austin, Austin Plumber, Plumbing Austin TX, Best Plumber Austin. This reads as spam to both Google and humans, and it tanks click-through.
What to do: Use your primary phrase once, naturally, and write the rest of the title for a person deciding whether to click. One clear, human title outperforms a keyword pile every time.
Part 2 — Meta Descriptions: Your Free Ad Copy
The meta description doesn't directly affect rankings — Google has said so plainly. But it's the snippet of text under your blue link, and it absolutely affects whether someone clicks. Treat it as the one-line ad you get to run for free.
4. Write a description for every important page
What "bad" looks like: No meta description at all, so Google grabs a random sentence from your page — often a navigation label or a cookie notice. Or the same generic description repeated site-wide.
What to do: Write a unique description for each page that earns its traffic. Summarize what the page delivers and include a reason to click. It's fine to skip descriptions on low-value pages; spend the effort where the searches are.
5. Aim for about 150–160 characters
Google truncates descriptions at roughly 155–160 characters on desktop (less on mobile). Anything past that gets cut.
What "bad" looks like: A 300-word paragraph dumped into the description field, or a 30-character fragment that wastes the space.
What to do: Write 150–160 characters that include your main keyword (Google bolds matching terms in the snippet) and a clear value proposition. Think of it as the subhead under your headline.
6. Don't expect the description you wrote to always show up
Here's the part that surprises people: Google rewrites meta descriptions more than half the time, pulling a snippet from the page body when it thinks that better matches the query. You can't force your description to display.
What to do: Write a good one anyway — it's your best shot at controlling the snippet, and Google uses it as the default. But don't lose sleep when Google substitutes its own. Focus your energy on the title, which Google respects far more often.
Part 3 — Headings: Structure Google (and Screen Readers) Can Read
Your headings (H1 through H6) are the outline of your page. Used well, they tell Google how your content is organized and make the page scannable. Used badly — or used purely for visual styling — they're noise.
7. Give every page exactly one H1, and make it meaningful
What "bad" looks like: No H1 at all (common when a CMS styles the page title as a plain <div>). An H1 that says "Welcome" or "Home." Or an H1 that's actually your logo image with no alt text.
What to do: Each page should have one H1 that states what the page is about — usually close to your title tag, but it can be more conversational. On a service page, the H1 might be Emergency Drain Cleaning, Same Day in Austin. One H1 per page is the safe, conventional rule.
8. Don't skip heading levels
Headings should nest like an outline: H1, then H2s under it, then H3s under those. Skipping from H1 straight to H4 because H4 "looked the right size" breaks the structure.
What "bad" looks like: Headings chosen for font size rather than hierarchy — an H3 followed by an H1 followed by an H5, in no logical order.
What to do: Use heading levels for structure and CSS for size. If you want a smaller heading, style it smaller — don't jump to a deeper level. This matters for SEO and it matters even more for screen-reader users navigating by heading, which is also covered in the website accessibility checklist.
9. Don't bury your headings in styling div soup
What "bad" looks like: Page builders that render every "heading" as a styled <span> or <div>, so the page visually has headings but the HTML has none. Google sees a flat wall of text.
What to do: View source and confirm your headings are real <h1>–<h3> tags. If your page builder fakes them, switch those blocks to true heading elements.
Part 4 — Alt Text, Canonicals, and URLs
These three are quieter, but each one causes real problems when it's wrong — and each is a quick fix once you know to look.
10. Give meaningful images descriptive alt text
Alt text describes an image for screen readers and for Google (which can't "see" images). It's also what shows if the image fails to load.
What "bad" looks like: alt="" on a product photo, or alt="IMG_4032.jpg", or alt="image" repeated across the page. Worse: stuffing alt text with keywords ("plumber austin plumber cheap plumber").
What to do: Write alt text that plainly describes what's in the image, as you'd describe it to someone on the phone: alt="Technician clearing a kitchen sink drain". Purely decorative images (dividers, background flourishes) should have empty alt="" so screen readers skip them — that's correct, not a bug.
11. Check your canonical tags point somewhere sensible
A canonical tag (<link rel="canonical" href="...">) tells Google which version of a page is the "real" one when duplicates exist. Most pages should canonicalize to themselves.
What "bad" looks like: Every page on the site canonicalizing to the homepage (a classic plugin misconfiguration that tells Google to ignore all your other pages). Or a canonical pointing to http:// when your site is https://, or to a staging URL.
What to do: View source and check the canonical on a few pages. Each should point to that page's own clean, preferred URL. If they all point to the homepage or somewhere odd, that's a settings problem in your SEO plugin worth fixing immediately.
12. Keep URLs short, readable, and stable
What "bad" looks like: /index.php?p=4471&cat=12&ref=nav, or /our-services/residential/plumbing/drains/emergency-drain-cleaning-services-in-austin-texas-area/. Long, parameter-heavy, or deeply nested URLs are harder for users and crawlers.
What to do: Use short, lowercase, hyphenated URLs that describe the page: /austin-drain-cleaning. Don't change URLs casually once they're live and ranking — if you must, set up a 301 redirect from the old URL so you don't lose the equity.
Part 5 — Thin Content and Cross-Page Duplication
The issues so far live on a single page. These last two only show up when you look across the whole site — which is exactly why they go unnoticed for years.
13. Find your thin pages
"Thin content" is a page with too little substance to be useful — a near-empty page Google has little reason to rank.
What "bad" looks like: Service or location pages with two sentences and a phone number. Tag/category archive pages that exist but hold one item. Placeholder pages that shipped and were forgotten.
What to do: There's no magic word count, but pages under ~250–300 words of real content deserve a look. Either build them out so they genuinely answer the searcher's question, consolidate several thin pages into one strong page, or remove the ones that serve no purpose. If you're not sure a page is pulling its weight, the content effectiveness guide walks through how to judge it.
14. Hunt down duplicate titles and meta descriptions
This is the big one. When ten pages share the same title tag, Google can't tell them apart, and they compete with each other for the same searches.
What "bad" looks like: A product catalog where every variant page is titled "Product | Acme Store." A location-pages template that stamped the same title on all 30 city pages. Pagination ("Blog — Page 2," "Blog — Page 3") that Google treats as near-duplicates.
What to do: You need to see all your titles and descriptions in one list to catch this. A crawler like Screaming Frog (free up to 500 URLs) will export every page's title and meta description so you can sort for duplicates in seconds. Then differentiate them — make each title and description specific to its page. This single fix often moves more pages than any other item on this list.
This checklist covers the on-page elements you can see and edit in your HTML. It deliberately stops at the page boundary. A few things it won't surface:
- Site-wide duplication at scale — you can spot-check a handful of pages by hand, but finding every duplicate title across a few hundred URLs needs a crawler.
- Whether Google has actually indexed a page — that's a Search Console question, covered in the DIY SEO Audit Checklist.
- Rankings and keyword gaps — on-page hygiene helps you rank, but it won't tell you which keywords you're missing.
- Performance, schema, and social previews — these are in your HTML too, but they each deserve their own pass (see the rich results and Core Web Vitals guides).
The honest limit: doing this by hand is fine for one page. Across a whole site, you'll miss things — not because the checklist is wrong, but because nobody manually compares 40 title tags without their eyes glazing over.
If you'd rather not crawl your own site and eyeball every tag, that's exactly what the On-Page SEO Audit ($39.99, one-time) is for. It crawls your homepage plus your internal pages, checks every one for title-tag quality, meta descriptions, H1 presence, heading hierarchy, image alt coverage, canonical tags, word count, and — the part that's tedious to do by hand — cross-page duplication. You get a prioritized, plain-English report of exactly what to fix and where, usually in under 24 hours.
It's a fraction of an agency retainer, with no subscription and no account access required. You point us at your domain; we hand you the punch list. Run the on-page audit and get your fix list without spending an afternoon in view-source.
Rather have it done for you? The On-Page SEO Audit pulls the data, runs every check above, and hands you a prioritized, plain-English report in under 24 hours — a fraction of an agency retainer, no subscription, no upsell.