SEO and AI search questions, answered

Publish the site on a public URL, add it to Google Search Console, submit a sitemap and ask Google to index your home page with the URL Inspection tool. Then get at least a few links from other sites, because Google finds most pages by following links. Google says a new site can take a few weeks to be noticed, so check back rather than resubmitting every day.

Usually one of five things: the site is too new and Google has not found it yet, a noindex tag or robots.txt rule is blocking it, no other site links to it, the pages have too little unique text, or it shows up but ranks too low to notice. Open Search Console's URL Inspection tool, paste your home page address, and it will tell you which of these applies.

Expect weeks for Google to find and index a new site, and months before it brings meaningful traffic. Google's own starter guide says some changes take effect in a few hours and others take several months. Fixes on an established site show fastest; a brand-new domain targeting competitive searches is slowest, because it has to earn links and trust that older sites already have.

On launch day, set up Search Console, submit a sitemap and make sure every page is crawlable with a unique title and description. Then pick specific searches you can win, such as questions, local terms and comparisons, and write one thorough page for each. Link those pages together in clusters, and earn your first links from profiles, directories, partners and communities. Expect results over months.

Yes, but not because Google judges whether a site is pretty. Design affects SEO through things Google can measure or read: whether text is real HTML, how pages are structured and linked, how fast and stable they load, whether they work on phones, and whether visitors find what they came for. A beautiful site that hides text in images or loads slowly can rank worse than a plain one.

If you want customers to find you through ChatGPT, Claude or Google's AI features, do not block their search crawlers. You can block training crawlers separately: OpenAI and Anthropic use different bots for training and search, and Google-Extended controls Gemini training without affecting Google Search. Blocking everything protects content from reuse but makes your site invisible in AI answers.

Not for Google rankings or AI Overviews: Google has said it does not use llms.txt, and that normal SEO is what gets content into its AI features. Beyond Google, support is uneven. Some AI coding tools and agents read llms.txt files, mainly for documentation sites, but no major AI search engine has said it relies on one. It takes minutes to add and does no harm, so treat it as a small extra, not a strategy.

Be indexed, be eligible for a snippet and be the clearest answer. Google says there are no extra requirements or special optimization for AI Overviews: they draw on pages in its index that can show a snippet. In practice, pages get cited when they answer a specific question directly near the top, back it with concrete facts and rank reasonably well for related searches. Do not block snippets with nosnippet or max-snippet if you want to appear.