5 SEO Basics That Still Move Rankings in 2026

Every year someone announces that on-page SEO is dead, and every year the same five checks keep turning up as the reason a page underperforms. Not because they're clever — because they're unfinished. A title that truncates, a page with three H1s, an image with no alt text: none of these are strategy problems. They're maintenance problems, and they're the ones that quietly cap what the rest of your work can achieve.

This is the audit I'd run before touching anything else. Every check has a threshold you can verify yourself in about ten minutes, and a fix you can ship the same day.

The short version

If you only read the table, read this one. These are the five checks, the thresholds that actually matter, and how to verify each without any tooling.

Now the reasoning — because the thresholds are useless if you don't know which way it's safe to break them.

1. Title tags: the strongest signal you fully control

Your title tag is the one ranking input you can change in thirty seconds and that a search engine will honour almost literally. It's also the one most people fill in once, at launch, and never look at again.

The length rule is really a pixel rule

"Keep it under 60 characters" is a proxy. Google truncates by pixel width, not character count — roughly 580px on desktop. A title made of narrow letters survives past 60 characters comfortably; one full of capitals gets cut at 50. Sixty is a safe approximation, not a law.

What actually matters is that the important words survive the cut. Truncating at "The Complete Guide to Website Auditing for Small Busi…" costs you nothing. Truncating at "WebGrowthSignal | The Complete Guide to Website…" means you spent your entire visible title on your own brand name, which nobody searched for.

Front-load the term, then earn the click

Two rules, in this order:

1. The primary keyword goes near the front. Not because early words carry some algorithmic bonus, but because a scanning human reads the first three words and decides.
2. Then make it worth clicking. A title is a search result and an ad. "SEO Audit Checklist" is accurate and forgettable. "SEO Audit Checklist: 5 Checks That Actually Move Rankings" tells someone what they get.

Duplicate titles are the other half of this. If four pages share the title "Services", you've told a search engine those four pages are interchangeable — so it will pick one and ignore the rest.

2. One H1, and it should agree with the title

Every page gets exactly one <h1>, and it should say the same thing as the title tag without being a copy-paste of it.

Multiple H1s won't earn you a penalty — this is not a rule with a punishment attached. What it does is muddy the page's topic at exactly the moment a crawler is trying to work out what the page is about, and it breaks screen-reader navigation, which is a real cost to real users regardless of what it does for rankings.

The usual cause is a design system: a hero component ships with an <h1>, someone adds a second hero further down the page, and now there are two. The second most common cause is the opposite — a page whose visible heading is a styled <div> because the H1 "looked too big", leaving the page with no H1 at all.

Verify it in five seconds. Open devtools and run this in the console:

document.querySelectorAll("h1").length

If it isn't 1, you have work to do.

Then get the outline right

Below the H1, headings should nest without skipping: H2 for each major section, H3 for sub-points inside it. Jumping from H1 straight to H4 because H4 was the size you wanted is a styling decision leaking into your document structure. Set the size in CSS; keep the level semantic.

3. Meta descriptions don't rank you — they get you the click

People dismiss this check because they've read that meta descriptions aren't a ranking factor. That's true, and it's beside the point.

The meta description is your ad copy in the search results. It doesn't move your position directly, but it moves your click-through rate at the position you already hold — and capturing more of the clicks already on offer at position six is usually cheaper than fighting your way to position five.

Aim for 120–158 characters. Under 120 and you're leaving space unused; past ~158 and it truncates. Search engines will also freely ignore what you wrote and generate their own snippet from the page body — which isn't a failure, it's a hint that your body copy answers the query more directly than your description does.

Write it as a promise:

Bad: "We offer SEO services for businesses of all sizes. Contact us today to learn more about our range of solutions."

Good: "Run a full on-page SEO audit in 10 minutes — titles, H1s, alt text, internal links and indexability, each with its threshold and its fix."

The second tells someone what they get and roughly how long it takes. The first could be about literally anything.

4. Alt text: accessibility first, image search second

Descriptive alt text is the rare check that's genuinely two things at once — an accessibility requirement and a route into image search. Do it for the first reason and the second arrives for free.

The part that trips people up is the decorative image. If an image carries no information — a background flourish, a divider, an icon next to a label that already says the same thing — the correct markup is an empty alt:

<!-- decorative: a screen reader skips it -->
<img src="/flourish.svg" alt="">

<!-- informative: a screen reader announces it -->
<img src="/score.png" alt="Growth Score dashboard showing a score of 72">

alt="" is an active statement: this image is decorative, skip it. A missing alt attribute says nothing at all, so a screen reader falls back to reading the filename — and hero-final-v3-compressed.png is not information, it's noise.

Which is why "every image needs alt text" is subtly wrong advice. Every image needs an alt attribute; whether it has content depends on whether the image carries meaning.

And describe the image's purpose, not its pixels. For a photo that links to your pricing page, the useful alt text is "See pricing plans", not "Man smiling at laptop".

5. Internal links: the check nobody does

Of the five, this is the one I most often find completely untouched — and it has the most leverage, because it's the only check here that compounds.

Internal links do two jobs:

• They distribute authority. A link from a page that has earned external links passes some of that value onward. Your strongest pages are an asset you can spend on your weakest ones.
• They're how pages get found. A page with no internal links pointing at it is an orphan. It may sit in your sitemap, but nothing on your own site vouches for it, so it gets crawled rarely and ranked accordingly.

The rule I actually use

Every new page links out to at least two existing pages, and is linked from at least one. That's the whole rule. Applied consistently it means you never create an orphan, and every new page inherits a little authority instead of starting from zero.

Two details matter more than the count:

• Anchor text should describe the destination. "Click here" and "read more" tell a crawler nothing. "How we monitor competitors" tells it exactly what's on the other end.
• Depth matters. A page four clicks from the homepage gets crawled less often than one that's two clicks away. If something is important, shorten the path to it.

The trap that outranks all five

Here's the uncomfortable part. You can nail every check above and still rank for nothing, because none of it applies to a page that cannot be indexed.

Indexability is binary, invisible on the rendered page, and by far the most common cause of the "we did all the SEO and nothing happened" conversation. Check these before anything else:

The staging noindex is the classic. A team sets it so the staging site stays out of the index, the flag is environment-based, someone gets the environment check wrong, and it ships. The site looks perfect. It is invisible. Nobody notices for a month, because there's no error, no warning and no visible symptom — the traffic simply never arrives.

Check it right now. View source on your most important page and search for noindex. It takes ten seconds and it is the highest-expected-value ten seconds in this entire article.

What to do this week

In priority order, because the order is the point:

1. Confirm your top 5 pages are indexable. No stray noindex, canonical pointing at itself, not disallowed in robots.txt. Nothing else on this list matters until this is true.
2. Fix duplicate and truncated titles, starting with pages that already earn impressions — you're recovering clicks you've already paid for.
3. Run the H1 check on every template, not every page. Templates are where the bugs live; fixing one template fixes hundreds of pages.
4. Rewrite meta descriptions for pages with high impressions and low CTR. Search Console tells you which ones those are, for free.
5. Adopt the internal-linking rule on the next page you publish. Two out, one in, descriptive anchors.

None of this is glamorous and none of it will make a good conference talk. All of it works, and most of your competitors haven't done it either.

The reason it doesn't get done isn't that it's hard — it's that it's invisible. Nothing turns red when a title tag drifts to 71 characters, so it drifts, and stays drifted, for two years. That's the real problem: not learning the rules, but noticing when you've quietly stopped following them. It's the argument for continuous auditing over one-off audits — re-run the checks on a schedule and turn each failure into a specific task instead of another number on a chart, which is the same case against dashboards generally.

Once your own house is in order, the next question is what everyone else is doing — and that turns out to be a monitoring problem, not a research problem.