SEO Checklist for New Website: 8 Checks Before You Blame Yourself

Is It Your Fault? Run This SEO Checklist for a New Website First

The fear every owner has when they open Search Console is that they did something wrong. Usually they did not. Google's own page on how to get a site on Google points out that a new site is often missing for the plainest reason there is: Google has not crawled it yet, and nothing else on the web links to it. We agree, and we would add that the dashboard will not tell you which of the two it is.

Most checklists start before launch: pick a domain, install a plugin, write titles. This one starts after. Your site is live, it has pages, and Google is not showing them. Google's SEO Starter Guide says there is no guarantee any page gets indexed, and we read that as permission to stop guessing and start checking. Each check below takes a browser, and a few take one command you can paste into a terminal. None of them needs a tool you pay for.

Before You Start: What 'Not Indexed' Can Actually Mean

Search Console uses two phrases that sound alike and mean different things. "Discovered, currently not indexed" means Google knows the address and has not fetched it yet. "Crawled, currently not indexed" means Google fetched the page, read it, and chose not to keep it. The first is a queue. The second is a judgement. Our page on What each Search Console status means for your business shows every status with one instruction each, and the counts below are from the same reads.

Two things in that table matter for your afternoon. On both reads, zero pages had a canonical chosen by Google that differed from the one the page declared. Our site is bilingual, and we had worried Google was treating the French and English pages as duplicates. The data said no in one column. And the 64 pages Google had not fetched became 16 within eight days, with recrawls dated 2026-10-12 to 2026-10-15, without any of them changing.

A new site starts with a small crawl budget. Practical Ecommerce's note on crawl budgets reports that Google crawls a new site conservatively until it has reasons to come more often, and we agree: "not indexed yet" is sometimes nothing more than "not yet". The checks tell you whether yet is the right word.

The Afternoon, in Order: Checks 1 to 5 (Host, robots and Sitemap, Head Tags, Googlebot's View)

Check 1, one host. Type your bare domain and your http address. Both should land on one final address. Ours: clevrstart.com and http:// both answer 308 to https://www.clevrstart.com. One site, one door.

Check 2, robots.txt. Open yoursite.com/robots.txt. It should allow Google in and name your sitemap. Ours reads Allow: / and Sitemap: https://www.clevrstart.com/sitemap.xml. Nothing in our robots.txt was wrong.

Check 3, the sitemap. Open the sitemap address. Every page listed should load with a 200. Ours: 105 pages, all 200. Then look for a date on each entry, the lastmod. Only 31 of our 105 carried one on 2026-10-14. A sitemap with no dates gives Google no reason to think anything changed. Google's Search Central documentation shows the sitemap format and the lastmod field, and the team there has said a wrong lastmod is worse than none, so put a true one or leave it out.

Check 4, the head tags. View the source of one page per template and look for rel="canonical". It should be the page's own full address. If you serve two languages, look for hreflang lines naming each language and an x-default. Ours carry en, fr-CA and x-default on every pair. Four pages did not; that story is in the next check.

Check 5, what Googlebot sees. Fetch a page as a browser and again as Googlebot and compare. Ours were identical: 200, the text in the HTML itself, 2,398 words on the English home page and 2,609 on the French, no noindex header.

Checks 6 to 8 and the One Defect: Old Sitemaps, Internal Links, Subdomains, and the Legal Pages

Check 6, old sitemaps. If your site was rebuilt or moved, the old sitemap addresses may still be registered in Search Console. Ours, /sitemap-index.xml, /fr/sitemap.xml and /sitemap-0.xml, all answer 404 and still show up in our analytics as requests. Siteimprove's list of redesign mistakes reports that a stale sitemap after a redesign is one of the most common errors, and Webflow's migration checklist shows resubmitting the sitemap as a launch-day step; we agree with both, and we would add: remove the dead ones too.

Check 7, internal links. Open your home page and your main pages and count the links to the pages that sell. On 2026-10-15 our home page, /tools, /info, /about and /work carried zero links to any service page. Only the services catalogue linked them, twice each, and the pricing page once. The five English service pages at the end of that graph, /services/ai-agent, /chatbot, /booking-software, /blog and /professional-website, had been "Crawled, currently not indexed" since 2026-08-28 or 29, while their French twins were indexed. We will not claim the missing links caused the stall; we report that the five pages nothing linked to were the five that stayed out. The fix was a Services list in the footer of every page, in both languages, commit 694d211, live the same hour. Our page on custom blog development is one of the five, which is why the data on it is ours.

Check 8, subdomains. Some parts of a site should stay out of Google. Our bookings and agents subdomains carry a noindex tag on purpose. They show up in the dashboard as "excluded by noindex", which is correct, not a problem.

The one defect. Four legal pages, privacy and terms in both languages, emitted a relative canonical, /privacy instead of the full address, and no hreflang. Found on 2026-10-14, fixed in commit 605cba5, verified live the next day. Nothing in the dashboard pointed at it. DesignRush's roundup reports that John Mueller has said crawled-not-indexed can sometimes reflect site quality; we take that seriously, and on our site the measured cause was links, not words.

Dashboard Versus API: Why the Number You See Is Not the Number to Act On

We opened our Search Console dashboard on 2026-10-15 and it said 82 pages were not indexed. The same afternoon, a script that asks Google's URL Inspection API about each of the 105 pages in our sitemap said 25. Both are true.

The 57 in between were not pages we wanted ranked. bookings.clevrstart.com answers with noindex on purpose. agents.clevrstart.com, the same. Two admin subdomains redirect to a login. The bare domain and http redirect to www. The dashboard counts all of that, and it refreshes every few days; the API answers for one page, now. The method behind the per-page read, from the first API call to the request, is what how to get google to index your site shows step by step, with the same tables.

One other site of ours tells the opposite story, and we include it so you do not take one case as a law. flowingkhaos.com had 1 of 26 pages indexed at 22:13 UTC on 2026-10-07 and 24 of 26 at 00:22 UTC the next day. Google crawled nothing in those six hours. The sitemap had named every page with a trailing slash while Google held the version without, so every entry was filed as a duplicate. Fixing the sitemap fixed the count. That is one site's case, and it is why check 3 comes before any theory. The structure a local business needs so the pages exist to be indexed at all is what our Local SEO Foundation service builds: one page per service, in the languages your customers use.

The One Thing to Do With a Page Still Not Indexed, and What We Don't Know Yet

Run URL Inspection on the page, in Search Console, one address at a time. Read the verdict. If it says "Crawled, currently not indexed", change something on the page or on what links to it before you ask Google to look again. Validate Fix re-checks the same judgement and returns the same answer when nothing changed. Request Indexing asks for a fresh look at the page as it is now, which only helps if it is not the same page. The order, from the status to the click, is what request indexing google shows, with the same reads and one rule: request only what you changed.

Google's page on getting a site on Google, linked above, says a page needs links and time. We agree, and we are honest about the time. We expect a page with nothing broken to be indexed within about 48 hours of a request. That is our experience from other sites, not a measurement from this one. Our five English service pages are the live test: links added on 2026-10-15, request not yet made as of that evening, and the dates will be published as they land. The Validate Fix click on 77 pages around 2026-10-07 is approximate, because we did not write the time down. The two to five weeks we expect before real rankings is an expectation. Google's starter guide, also linked above, says plainly that nothing guarantees indexing, and the checklist cannot change that; it can only remove every reason Google has to say no.

Next Step: Read the Local SEO Plan, Then Decide Whether to Hire Anyone

The checklist is an afternoon's work, and most owners can do it alone. Do it before you pay anyone. If every row is a pass and the pages still sit in "discovered", give Google links and dates and a week. If a row fails, you have found the thing that was never your fault and is now your fix.

If you would rather have the pages built and linked right the first time, the Local SEO plan shows what we build: one page per service and per area that earns it, in English, French or both, with your Business Profile set up to match and a report a month. Every button on this site books through the same form, so the next step is a call, not a cart. According to our services page, which lists what each one includes with its scope, every service is quoted after an evaluation, and we agree that is the right order: check first, then quote.

A status is a signal to read, not a verdict on your business. Measure before you fix. Do it this week. Then tell us what broke. Probably nothing.