The on-page SEO checklist I run on every page that matters

This is the checklist I run before I quote for on-page SEO work, and I'd rather you ran it yourself first. It takes around an hour for the first page you run it on, it needs nothing you have to pay for, and it will tell you whether the page has a targeting problem, a writing problem or no problem at all. A fair share of the pages I'm asked to look at fail on the first two checks, and the title fix at least is one an in-house marketer can make in a morning without paying anyone.

The order matters. It runs from the things that decide whether a page can rank at all down to the things that nudge a page already in contention. If you only have twenty minutes, do the first three and stop.

1. Does the page match what the searcher wants?

Search the keyword yourself, in a private window so your own history doesn't skew it, and look at what Google has chosen to show. If the top ten results are all guides and your page is a service page, or they're all product listings and yours is a 2,000 word explainer, the page is aimed at the wrong search and no amount of polishing fixes that.

The second half of this check is whether the page has one job. Open Search Console, go to Performance, filter by the page URL and look at the queries it gets impressions for. A page that shows up for "boiler repair guildford" and "how does a combi boiler work" is trying to be two pages, and it will lose to a competitor who wrote one page for each. Pick the query that pays, and send the rest to a page of their own. I've written up how that plays out on a shop in the category page SEO guide, where the category page and the buying guide need to be different URLs.

2. The title, and whether Google will keep it

The title is the line in the search results people decide to click or not, so it does two jobs at once. It has to carry the main phrase, near the front, and it has to read like something a person would choose over the nine results around it. "Boiler Repair Guildford | Same-Day Call-Outs" does both. "Home | Smith Heating Ltd" does neither, and it's still the most common title on the small business sites I audit.

Length matters more than it looks, because Google rewrites titles it doesn't like. Zyppy's study of 80,959 title tags across 2,370 sites found Google rewrote 61.6% of them, and the rewrite rate climbed with length. Titles over 70 characters were rewritten 99.9% of the time, while the 51 to 60 character range had the lowest rate at around 40%. The data is from early 2022, and I've seen nothing since to suggest the pattern has changed. Keep titles under 60 characters and you keep control of what appears on the results page.

3. The first screen and the headings

Read only the main heading and the subheadings on the page (the H1 and H2s, if you're looking at the code), nothing else, and ask whether you'd know what the page says. Headings like "Our Services", "Why Choose Us" and "Get In Touch" tell a reader and a search engine nothing. "What a boiler repair costs in Guildford", "What we fix on the first visit" and "When a repair isn't worth it" tell both exactly what's coming, so the headings should be the questions.

The first screen gets the same test. Whatever the searcher typed, the answer, or at least the shape of it, should be visible before they scroll. A service page that opens with three paragraphs about the company's history and values before mentioning what it does or where has pushed the answer below the fold on every phone. Give the page one main heading that matches the intent of the title without copying it word for word (the boiler page's title can be "Boiler Repair Guildford | Same-Day Call-Outs" and its heading "Same-day boiler repair across Guildford"), and let the opening paragraph answer the question before it does anything else.

4. Does the body answer the question better than the pages above it?

Open the three pages ranking above you and list the subtopics they cover that your page doesn't. That's the whole check, and it comes as a relief to anyone who has been told their page needs to be 2,000 words, because word count has nothing to do with it. If all three answer "how long does it take" and "what does it cost" and yours answers neither, that's the gap, and closing it is a far better use of an afternoon than adding 800 words of padding.

Thin content fails this check, but so does bloated content. A page that gives the answer in its first 300 words and then covers the follow-up questions (for the boiler page, what it costs, how long it takes and whether an old boiler is worth repairing at all) is in a far better position than one that takes 1,500 words to reach the answer, because the searcher gets what they came for and stays on the page.

5. The links into and out of the page

A page with no internal links pointing at it is invisible to Google until a sitemap or an external link rescues it, and it inherits none of the authority the rest of your site has built. In Search Console, open Links and look at "Top linked pages" under internal links. If a page you want to rank isn't in that list, or has fewer links than your privacy policy, that's the first thing to fix, and it costs nothing. Link to it from the pages that already rank, using the phrase you want it to rank for as the link text rather than "click here" or "read more".

It's worth flagging that links out matter too. Every guide on this site links up to the service page it supports and across to two or three related guides, which is how the technical SEO checklist and the pages around it reinforce each other rather than compete. If your page is a dead end, add the two or three links a reader would want next.

6. Speed and images

Run the page through PageSpeed Insights and look at the field data section, which is what real visitors experienced rather than a lab simulation. Google's thresholds for a "good" page are a Largest Contentful Paint (LCP) of 2.5 seconds or under, an Interaction to Next Paint (INP) of 200 milliseconds or under and a Cumulative Layout Shift (CLS) of 0.1 or under, each measured at the 75th percentile of visits. I've explained what those three mean in plain terms in the Core Web Vitals guide, but for the checklist the question is whether the page passes, and if it doesn't, whether images are the reason. On small business sites they usually are. A 4MB photo at the top of the page, exported straight from a phone, uncompressed and served at full size to a mobile screen, is the single most common cause of a failed LCP that I see.

While you're there, every image that carries meaning needs alt text describing what's in it, in plain words, without stuffing the keyword into every one.

7. The small things that still change clicks

These sit last because they rarely decide whether a page ranks, though they do decide what happens once it does. Google doesn't rank on the meta description, but it's the text under your title in the results, so it should say what the page offers and why to pick it, in around 150 characters, and Google will write its own if yours is missing or vague. The URL should be readable and short, and if you change it, the old one needs a redirect, which I've covered in the redirects guide because a renamed URL with no redirect throws away everything the page had earned. Structured data, the code added to a page that tells Google what kind of thing the page is, belongs here too, and the schema markup guide sets out which types are worth adding and which are a waste of a developer's afternoon.

What's deliberately not on this list

A few items turn up on other on-page checklists that I leave out on purpose, and it's worth saying why.

• Keyword density, because there's no percentage to hit and chasing one produces copy nobody wants to read.
• Minimum word counts, for the reason in check four.
• "LSI keywords", a term that describes nothing Google uses.
• The exact keyword in the alt text of every image, an old trick that now reads as spam.
• Meta keywords, which Google stopped reading in 2009.

The checklist in one place

Check	What good looks like	Where to look
1. Intent match	Your page is the same type as the top ten results and targets one query	A private-window search, plus Search Console Performance filtered by page
2. Title	Main phrase near the front, under 60 characters, reads like a reason to click	The page source, or the Search Console query report for click-through rate
3. Headings and first screen	The main heading and subheadings read as the questions the searcher asks; the answer starts before the first scroll	The page on a phone
4. Body	Covers the subtopics the top three pages cover, then stops	The three pages ranking above you
5. Internal links	Linked from pages that already rank, with descriptive link text; links out to the next step	Search Console Links report
6. Speed and images	LCP under 2.5s, INP under 200ms, CLS under 0.1 in field data; images compressed and sized	PageSpeed Insights
7. Meta description, URL, schema	Description written, URL readable and redirected if changed, schema that matches the page	The page source and the Rich Results Test

If a page passes all seven and still isn't ranking, the problem usually sits outside the page, in the site's authority or its technical foundations, and that's where the technical SEO checklist picks up.

On-page SEO checklist FAQs

How long does it take to run this on a page?

Around an hour for the first page, because checks one and four involve reading competitors, and twenty minutes or so for each page after that once you know the pattern. Running it across a whole site is a different job, and at that point it's worth prioritising the pages that already get impressions in Search Console but sit on page two, since those move fastest.

Do I need any paid tools?

No. Search Console, PageSpeed Insights and the Rich Results Test are free, and a private browser window covers the search-results check in step one. Paid tools make the competitor comparison in check four quicker across a large site, but they don't change the verdict on any single page.

How soon do on-page changes show up in rankings?

In my experience, changes to pages Google already crawls regularly show within two to four weeks, and a title change can register within days. Pages that have never had impressions take longer, because Google has less data to re-evaluate. Either way, note the date you made the change and compare the page's queries in Search Console a month later rather than refreshing the rankings daily.

What if the page fails the first check?

Then don't spend time on the other six until you've fixed it. Either rewrite the page for the query it can realistically own, or build a second page for the intent it's currently half-serving. It's also the first thing I check when someone asks me to look at their pages.