Short answer
What should a DIY SEO checklist include?
A DIY SEO checklist should cover five things, in this order: whether search engines can index your pages, what each page says it is, whether your links work, whether the content is worth ranking, and a few technical extras. The 25 checks below can all be done yourself with free tools: Google Search Console, a crawl of your site and your own reading.
They are numbered in the order to do them. Each one says what to look for and where to look, and the last section is honest about the few that need more than free tools.
How to use this checklist
Start with a baseline: a scan here of up to 5,000 pages, and a note of the clicks and impressions Search Console reports for the last three months. The scan answers most of the mechanical checks at once and groups its findings by cause, so one faulty template is one task rather than a hundred.
Then work through the sections in order. Indexing comes first, because a page that cannot be indexed gains nothing from a better title. Within each section, start with the pages that already get impressions. Re-scan after each round of fixes to confirm the faults are gone and nothing new has broken.
Indexing and crawlability checks
Each of these decides whether a page can appear in search at all.
- Your important pages are indexed. In Search Console, open Pages under Indexing and read why any excluded pages are excluded. URL Inspection checks a single address.
- No stray
noindex. One left over from a staging site or set by a plugin removes a page from search. A scan's noindex check lists every page that carries one. - robots.txt blocks only what you mean it to. Look for
Disallow: /or a rule that catches a whole section by accident. The robots.txt tester shows how each rule resolves. - The XML sitemap lists only pages you want indexed. Every URL in it should answer 200 and be the canonical version, and the sitemap should be submitted in Search Console. The sitemap validator checks the file.
- Canonical tags point where they should. An ordinary page's canonical points at itself; a copy points at the original. A canonical aimed at the home page or a deleted URL quietly drops a page from search. The canonical checker reads one page's for you.
- The site has one address. The http, https, www and non-www versions should all redirect to one of them, in a single hop.
- The content is in the HTML. If text or links only appear after JavaScript runs, crawlers that do not run scripts miss them. A scan with “Use browser” on compares the HTML your server sends with what a browser builds.
On-page checks: titles, descriptions, headings
- Every page has a title. Missing titles are among the first things a scan flags; the bulk meta tag checker lists titles and descriptions for the whole site.
- Titles are unique and specific. Each says what the page is, in the words a customer would search, with the important words first. Two pages with the same title are either competing or one is a duplicate.
- Meta descriptions describe the page. Google often rewrites them, but a clear one is what shows when it does not. A template that leaves them blank or identical is one fix for the whole site.
- One H1 per page, saying the same thing as the title in a reader's words. The H1 checker finds pages with none or several.
- Headings describe the sections under them and run in order, rather than being used for large text.
- The answer is in the first paragraph. What the page offers, the price, the place, before any background.
Link checks: broken, internal, external, redirects
- No broken internal links. Fix each link where it is written rather than relying on a redirect. The broken link checker lists them with the page they are on.
- Internal links go straight to the final address. A link through a redirect, and worse a chain of them, slows visitors and crawlers; point it at the destination.
- Every important page is linked from other pages. A page nothing links to, an orphan page, is hard to discover. Keep the pages that earn money within a few clicks of the home page.
- Link text says where the link goes. “Boiler servicing prices” tells readers and search engines more than “click here”.
- Outbound links still work. Pages you cite move and close, and a dead outbound link is a sign nobody maintains the page. The external link checker asks each destination whether it still answers.
Content checks: duplicates, thin pages, spelling
- No two pages chase the same search. Merge them into the better page and redirect the other address to it. The near-duplicate check lists pages that say almost the same thing.
- Thin pages are improved, merged or removed. Empty tag and category pages, near-identical town pages. Ask whether anyone landing on the page from a search would be glad they did.
- Facts are correct and spelled right. Prices, opening hours, phone numbers, product names and dates. Read your key pages through yourself; no crawler knows what your prices should be. For spelling, a scan spell-checks the whole site.
- Images are a sensible size. A photo straight off a phone is far larger than any page needs. Resize it to the size it is shown at and compress it.
Technical extras: hreflang, schema, accessibility
Only some of these apply to your site. Skip the ones that do not.
- hreflang, if you publish in more than one language or for more than one country. Each version lists every version, itself included, and the references must go both ways; a one-way reference is ignored.
- Structured data is valid. If you mark up your business, products or FAQs, test the pages in Google's Rich Results Test and make sure the markup says what the visible page says. To find which pages have a problem first, the bulk schema validator checks every page at once.
- Accessibility basics. Alt text on images that carry meaning, link text that makes sense on its own, enough contrast. A scan with “Use browser” on runs axe-core, the engine behind Lighthouse's accessibility audit, on every page of the site; Lighthouse in Chrome's developer tools does it for one page. Going through your main pages with only the keyboard catches some of what both miss.
Which checks need a paid tool or an expert
None of the 25 needs a paid tool. Paid tools add data about other people's sites: backlink indexes, search volumes for queries you do not rank for yet, and competitor rankings. That is useful for research and not needed to finish this list. Doing SEO for free goes through what each free tool covers.
Some checks need an expert's judgement on some sites:
- Moving domains or platforms. Every old address needs a redirect, and getting that map right takes experience.
- Very large sites. Tens of thousands of pages or faceted navigation turn indexing and canonical decisions into engineering.
- Content built by JavaScript that a scan shows crawlers are missing.
- hreflang across many countries and languages.
- A sudden fall in traffic with no obvious cause. Check the Manual actions report in Search Console first.
There is also a size limit: a scan here reads up to 5,000 pages. On a larger site, use path filters to scan one section at a time.
Questions people ask
- How often should I go through a DIY SEO checklist?
- Once in full, then in parts. Re-scan after anything structural, such as a redesign, a new section or renamed pages, because that is when indexing and link faults appear. Look at Search Console once a month and go through the whole list again every few months. A free account here can schedule repeat scans, so the crawl side happens without you having to remember it.
- Do I have to fix every item before anything improves?
- No. The indexing checks come first because they are all-or-nothing: a page that cannot be indexed earns nothing however good its title is. After those, start with the pages that already get impressions in Search Console. The 80/20 rule in SEO explains how to choose that order.
- Can a site crawler do the whole checklist for me?
- Not all of it. A crawler is good at the mechanical checks: broken links, redirects, missing or duplicated titles, pages carrying a
noindex. It cannot tell whether a page answers the question people searched or whether your prices are current, and it cannot tell you which pages Google has actually indexed. For that you need Search Console. - Is a checklist enough to rank?
- It removes what stops you ranking; it does not make a page worth ranking. That still takes pages that answer real searches better than the alternatives, and other sites that link to them. Doing SEO on your own covers the rest of the job.
Run the mechanical checks in one go
Read next
- The DIY SEO guide — the whole job, step by step, in the order it pays off.
- How do I do a DIY SEO audit? — five steps in order, from a free crawl and indexing checks to links, titles and a re-check.
- Is it possible to do SEO on my own? — what the work is, what it takes, and where doing it yourself runs out.
- Can I do SEO myself for free? — the free toolkit, a first week of work, and what paid tools actually add.