Technical SEO checklist: 35 checks for Google and AI search (2026)

A technical SEO checklist tests whether search engines and AI crawlers can reach, render, index and understand your pages. Run it in this order: crawl access, indexing, rendering, speed, site structure, structured data, then measurement. The order matters. A fast page that Google never indexes earns nothing, so a failure early in the list makes every fix after it worthless. Most audits find a handful of faults that cost real money and hundreds of warnings that don't. Your job is to find the handful.

I'm Brenden House. I run Searchmaxxed, an SEO and AI SEO agency, and on 27 October 2026 (AEST) I checked every limit below against Google and OpenAI documentation.

In summary

• Seven areas, 35 checks, in order. Crawl access, indexing, rendering, speed, structure, structured data and measurement, with where to test each check and what a pass looks like.
• Fix in revenue order, not tool order. A wrong canonical on your pricing page matters more than 400 missing alt tags on old blog posts.
• JavaScript sites fail quietly. Googlebot reads only the first 2MB of an HTML file, and links need real <a href> tags, two things single-page apps often break.
• AI search needs nothing exotic. Google says AI Overviews and AI Mode have "no additional requirements", so let crawlers in and keep OpenAI's search and training crawlers apart.
• Check your tracking hard. GA4 won't count a key event whose name runs past 40 characters, and a crawler won't catch it.

What is a technical SEO checklist?

A technical SEO checklist is a list of tests for the parts of a website that decide whether search engines can use it at all. It covers servers, code, URLs and site settings, not the words on the page. It's one of four kinds of SEO:

Two terms come up constantly:

• Crawlability is whether a crawler can fetch a URL, which depends on robots.txt, server responses and the links that lead to it.
• Indexability is whether a search engine is allowed to store a page and show it in results, which noindex tags and canonicals control.

A technical SEO audit runs this checklist across a whole site and ranks every fault by cost. If you're weighing what a full audit should include beyond the technical layer, our guide to what an SEO audit covers sets it out.

Can search engines and AI crawlers reach your pages?

Crawl access is the first gate. If robots.txt, a firewall or a server error stops Googlebot or OpenAI's crawler, nothing else on this list counts. Start here.

Google caps robots.txt at 500 KiB and ignores anything after that, according to its robots.txt documentation. Most sites never get close. Large ecommerce and documentation sites sometimes do.

Check 2 catches more enterprise sites than people expect: a security team turns on bot protection, and crawlers get a challenge page instead of content. Google's own AI features guidance lists "Ensuring that crawling is allowed in robots.txt, and by any CDN or hosting infrastructure" first among the fundamentals.

Check 6 is a choice. OpenAI's crawler page says sites that opt out of OAI-SearchBot "will not be shown in ChatGPT search answers". GPTBot is the separate crawler for training data, so you can allow the first and block the second. OpenAI says changes take about 24 hours.

Are the right pages indexed?

Indexing checks confirm that Google stores the version of each page you want ranked. The Page indexing report and the URL Inspection tool in Google Search Console answer most of them.

Google's sitemap guide limits each sitemap to "50MB (uncompressed) or 50,000 URLs", so bigger sites split them and submit a sitemap index.

Two Pages report statuses matter most. "Discovered, currently not indexed" means Google knows the URL but hasn't fetched it yet. "Crawled, currently not indexed" means it fetched the page and chose not to store it. The first usually points to crawl capacity or weak internal links, while the second usually points to a problem with the page itself.

How fast can a clean site get indexed? Here's a site we own:

We launched a new site with 161 guides, each built for one search from the live results and People Also Ask. We submitted the sitemap and pinged IndexNow. No paid indexer. About 25 hours after Google's first crawl, 203 of 228 URLs were indexed. Indexed isn't ranked, though: 5,198 impressions in ten days and 5 clicks.

That last line matters: technical SEO gets you into the race, and pages that answer the search better than the competition win it.

Does Google see what your users see?

Rendering checks confirm that the HTML Google processes holds your content and links. This is where B2B SaaS and modern marketing sites built on React, Vue or Angular most often fail without anyone noticing.

Most checklists miss check 16. Google's Googlebot page, last updated 3 February 2026, says Googlebot "crawls the first 2MB of a supported file type". Once it hits the cutoff, it "stops the fetch and only sends the already downloaded part of the file for indexing consideration". The limit applies to the uncompressed file. A pricing page with a huge inlined JSON blob can push its real pricing table past that line, and Google never sees it.

Check 15 trips up design systems. Google's JavaScript SEO guide says Google "can only discover your links if they are <a> HTML elements with an href attribute". A card that navigates with an onClick handler works for a user. To a crawler, it's a dead end.

When we build a site, pages share templates that handle rendering, speed, indexing controls and schema, so we fix a rendering fault once. It stays fixed on every page.

Is the site fast enough on mobile?

Speed checks test Core Web Vitals, Google's three measures of loading, responsiveness and layout stability, on real users' phones. Google indexes mainly the mobile version of most sites, so test mobile first.

The thresholds come from Google's Core Web Vitals page, and Google's Googlebot page says that for most sites it "primarily indexes the mobile version". INP is the responsiveness metric, measuring how long the page takes to react after someone taps or clicks.

Decide on field data. Field data means measurements from real visitors, while a lab score in PageSpeed Insights is a test run on one simulated device. A lab score helps you debug, and it can still pass while real users on slow phones fail.

Does your site structure lead crawlers to your money pages?

Structure checks confirm that internal links send crawlers, and the authority links carry, to the pages that make you money. A good structure also tells Google which pages matter most.

The three-click line is my working rule, not a Google number. It keeps a big site honest. Orphan pages are the quieter problem, because Google finds them through the sitemap, if at all, and rarely treats them as important.

Links are strategy too. On the sites we build, each page has a distinct job, and we connect related pages through internal links as they ship.

Do search engines and AI answers understand what each page is about?

These checks confirm that structured data, titles and language tags describe each page accurately, which matters for rich results and AI answers.

Structured data, or schema markup, is code that labels facts on a page in a format machines read. Think product name, price or author. JSON-LD, a script block that sits apart from the visible HTML, is the format Google recommends.

Selling in Australia and the US? Then check 32 matters. Google's guide to localised versions says "Each language version must list itself as well as all other language versions". It names missing return links as a common mistake. When /en-au/pricing points to /en-us/pricing but the US page doesn't point back, Google may ignore the tags. If you run several markets, our international SEO service covers the setup.

Does AI search need anything extra here? Google's AI features guidance answers it: "There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." Crawlable pages, clear facts and matching schema do the work. Getting named in ChatGPT and Perplexity also depends on what other sites say about you, which sits outside a technical audit. Our AI search source readiness checklist covers that side.

Can you measure what each fix changed?

Measurement checks confirm that Search Console, Bing Webmaster Tools and GA4 record what your fixes do. Without them, every fix is a guess.

Check 34 comes from a real account. Google's GA4 limits page says that if a key event's name "exceeds 40 characters, then the event will not be reported as a key event". On an enterprise B2B client's account:

A client's four enquiry types never counted as conversions. Their event names ran 41 to 50 characters; GA4 cuts at 40. The 40-character "career" event counted every time. Nobody had noticed, because the events still showed up.

Same account, check 35:

A client's GA4 showed AI-assistant visits jump from 354 to 6,050 in a month. Almost none of it was real. About 5,700 came in three one-hour bursts: same 1920x1080 desktop Chrome, all new users, no referrer, about 20 seconds each, landing on two pages. Real AI traffic was 385 sessions against 354 the month before, roughly flat.

Neither fault shows up in a site crawler, and both would have wrecked any report on whether the technical work paid off. So I treat measurement as part of the technical audit, and every Searchmaxxed managed package includes Search Console and GA4 wired and checked.

How do you run a technical SEO audit, step by step?

Run a technical SEO audit in seven steps, from collecting the data to verifying fixes on the live site. Size sets the timeline. A small marketing site can take a day or two, while a site with hundreds of thousands of URLs takes weeks.

1. Collect the data. Get Search Console, Bing Webmaster Tools and GA4 access, plus a list of the pages that bring revenue.
2. Crawl the site. Use a desktop or cloud crawler set to Googlebot Smartphone, with JavaScript rendering switched on.
3. Compare three URL lists. Put the crawl, the sitemaps and Search Console's indexed pages side by side, and the gaps show orphans, stray URLs and unindexed money pages.
4. Test rendering and speed by template. Pick one URL per template and run URL Inspection and PageSpeed Insights on each.
5. Read server logs on large sites. Logs show which URLs Googlebot fetches, how often, and which ones it never touches.
6. Rank faults by revenue. The priority table in the next section shows how to sort them by what each one costs.
7. Fix, then verify on the live site. Re-test each URL after release, and recheck indexing a week or two later.

Tools you'll want: Google Search Console, Bing Webmaster Tools, a crawler such as Screaming Frog or Sitebulb, PageSpeed Insights and the Rich Results Test. Sitebulb and Moz both publish free audit templates if you want a spreadsheet to work from.

Which technical SEO issues should you fix first?

Fix first whatever stops a revenue page from being crawled, indexed or tracked, and fix last whatever touches pages that earn nothing. A site audit tool's health score ranks issues by how often they appear, which is the wrong order.

A broken canonical on a money page outranks four hundred cosmetic warnings. That's how we rank work at Searchmaxxed: by what each fault is worth, not by how many times the crawler flags it. A tool can't do that part. It doesn't know which pages bring in pipeline.

What changes on enterprise and B2B SaaS sites?

Enterprise and SaaS sites need the same 35 checks, plus three things smaller sites can skip: crawl budget, release regressions and split ownership. Google's crawl budget guide applies to sites with "1 million+ unique pages" that change weekly, or "10,000+ unique pages" that change daily.

Below that, crawl budget is rarely your problem. Above it, faceted navigation and parameter URLs can eat Googlebot's time before it reaches new pages.

Release regressions are the bigger risk. One deploy can ship a noindex tag from staging, drop a canonical from a template, or swap links for buttons. On a large site that change hits thousands of URLs at once, so test the top templates after every release, not once a year.

A worked example: a B2B SaaS company selling in Australia and the US

Take a B2B SaaS company with a Next.js marketing site, docs on a subdomain, a blog on another CMS and the product on app.. Here's where the checklist usually finds trouble on a setup like that:

That last event name is an illustration, not a client's. Count your own. Each fault has a different owner, from web team to docs, product engineering and marketing ops, so technical SEO stalls in big companies. The audit is the easy part. Getting fixes shipped across four teams is the hard part, which is why we build the fixes ourselves and deliver them ready to merge. Our enterprise SEO service covers how.

How often should you run a technical SEO audit?

Run a full technical SEO audit once a quarter, and before and after any migration or redesign. Monitor the top checks weekly, because one bad release can undo a quarter's work.

My weekly short list:

• Search Console Pages report: any jump in "not indexed" on money templates
• Core Web Vitals report: any template that drops to "needs improvement"
• A crawl of the top templates after each release, with JavaScript rendering on, compared with last week's crawl
• GA4 key events: counts in line with the forms submitted

On the sites we manage, this runs inside the Managed Search Loop, our weekly cycle. Each week we measure, find the constraint, rank the fix, ship it, check the live page and record what changed.

Where does Searchmaxxed fit?

Searchmaxxed is an SEO and AI SEO agency for enterprise, B2B SaaS, venture-backed, manufacturing and public-sector companies. We do technical SEO, on-page SEO, content, link building, digital PR, GEO and AEO. The aim: rank in Google, and have ChatGPT, Gemini, Perplexity and Google's AI answers name you for the searches that bring pipeline.

On the technical side, that means the 35 checks above, run continuously. We rank every fault by the revenue behind it. Then we build the fixes ourselves, verify them on the live site, and our technical SEO service page sets out the detail.

Technical, schema and internal-link work doesn't reduce the page-release count in either of the first two packages. Every managed package includes ongoing audits, a 90-day plan in weekly releases, the build work itself and a monthly decision memo. You own the site, repository, content and data from day one. Want the audit on its own first? The AI SEO Audit takes five business days, shows what your site is doing to itself and ranks every fix by what it's worth.

Here's how we do it. We build on a website you own, our Agentic Website, with templates that handle rendering, speed, indexing controls and schema. AI agents pull your Search Console and GA4 data, read the Google results and check what AI answers say about you. That feeds the Managed Search Loop. It shows which fix or page should bring in the most enquiries next. We build it, check the live page and measure what changed.

Searchmaxxed's work has generated more than A$10 million in revenue through search for its clients.

Is Searchmaxxed right for you?

We suit marketing and growth teams at enterprise, B2B SaaS, venture-backed, manufacturing and public-sector companies who want technical fixes shipped, not listed. If you have an engineering team that ships SEO tickets fast, you may only need the checklist above.

In the right-hand column? A crawler and this checklist will serve you better than an agency.

Frequently asked questions

What should you do next?

Run checks 1, 7, 9, 15 and 34 on your five most valuable pages this week. They catch the costliest faults. If you'd like us to run the full checklist and ship the fixes, let's talk pricing.