SEO for a New Website: Launch Checklist and First Month

SEO for a new website comes down to two jobs. First, remove every reason Google can't find, crawl, or index your pages. Second, launch a small set of pages that each deserve to rank for one specific search. A brand-new domain has no links, no history, and no data. You can't win on authority yet, but you can control whether your site is technically findable and whether each page is the best answer to a narrow question.

This plan is for founders and small teams launching a first website. If you are replacing a site that already gets search traffic, use our website redesign SEO checklist instead. A redesign has existing rankings to protect, and that is a different job.

In this guide

What "zero" looks like, using our own site's dated numbers
Phase 1: seven things to build before launch, including how to pick searches a new domain can win
Phase 2: a launch-day checklist with pass conditions, and our own results from running it
Phase 3: what to watch and fix in weeks one to four
How long ranking takes, separated into three clocks

What zero actually looks like

We are a new site too, and our build log for this website records how it was made. Here is what Ahrefs reported for polydraft.ai on two dates, five weeks apart:

Ahrefs metric	August 29, 2026	October 5, 2026
Domain Rating	0.6	7
Referring domains	4	40
Organic keywords	0	0
Estimated organic traffic	0	0

Referring domains grew tenfold, and Ahrefs still tracked zero organic keywords. That doesn't prove links are useless, and third-party tools miss queries that Search Console would show. It does show that a tenfold jump in referring domains didn't produce any rankings Ahrefs could detect within five weeks. Pages also have to be indexed and be a precise match for a query.

It helps to separate what you control at launch from what you don't:

You control now	You don't control yet
Which pages exist and what each one is for	How many sites link to you
Whether pages can be crawled and indexed	How fast Google crawls and indexes them
Titles, headings, copy, and internal links	Which queries you rank for first
Whether each page adds something new	Whether Google shows your page at all

Google states plainly that it doesn't guarantee it will crawl, index, or serve a page, even one that follows its guidelines. Everything below removes obstacles. None of it guarantees a result.

Phase 1: before launch, build it findable

1. Decide which pages exist, one search intent per URL

Before you write any copy, list the searches you want each page to answer. Give each distinct search need exactly one page. Two pages chasing the same intent compete with each other. One page trying to cover five unrelated needs answers none of them well. Settle your URL pattern now too, because changing URLs after launch means redirects and a period of re-crawling.

A typical launch set for a startup or service business looks like this:

Page	Its job	Example search it could own
Homepage	Say what you are and who it's for	your brand name; your category plus audience
One page per core offer	Explain one product or service in depth	"bookkeeping for startups"
Prices	Answer the cost question directly	"[your category] cost"
About	Show who is behind the business and why to trust them	your brand name plus "founder" or "team"
Contact	Make the next step obvious	rarely a search target, still essential

If you have only one offer, a single page can work. Our guide to one-page website SEO explains when one page is enough. For larger sites, website architecture for SEO covers how to test which queries share an intent.

2. Pick searches a new domain can actually win

Broad terms are out of reach for a while. Choose queries where today's results show you have a chance. Our working rule: a query is worth targeting early when its intent matches one of your pages and the current top five already includes a weak result, such as a small site, a forum thread, or a page that only half answers the question.

Here is how that filter worked on our own keyword research. On October 5, 2026, we took a set of 167 English US queries that Ahrefs rated at a Keyword Difficulty of 20 or less. We then kept only those with at least 100 monthly searches and a site with a Domain Rating of 20 or less somewhere in the top five. Twelve queries were left. Several had nothing to do with our product, such as "words with geo in it", so checking intent cut the list further. A tool filter finds candidates; it doesn't pick them. Without a paid tool, you can do the same by hand: search the query, and look for forums, small sites, or off-target pages near the top.

Treat difficulty scores as estimates, not promises. After launch, Search Console's Performance report shows which queries Google already associates with your pages, and that becomes the better list.

3. Give every page something the current results lack

Google's guidance on helpful, people-first content asks whether a page provides "original information, reporting, research, or analysis." It also asks who created the content, how, and why. For a new business, original value is usually close at hand:

what you actually charge and what's included
how your process works, step by step
the questions customers really ask you
a constraint or trade-off you know from experience that generic guides gloss over

Don't launch with dozens of thin pages, such as one page per city with the city name swapped. Google's spam policies describe scaled content abuse as "when many pages are generated for the primary purpose of manipulating search rankings and not helping users." The policy applies however those pages were produced. Five strong pages beat fifty interchangeable ones.

4. Keep the staging site out of Google, then remember to undo it

Put your staging site behind a password. Google notes that robots.txt is not a mechanism for keeping a web page out of Google: a blocked URL can still be indexed if other pages link to it. A noindex rule is a reasonable backup on staging, but Google can only see it if the page isn't blocked by robots.txt.

A common and costly launch mistake is a staging noindex, or a Disallow: / in robots.txt, that ships to production. Visitors see a normal site, so nothing looks broken until someone checks. Phase 2 checks for exactly that, and our robots.txt examples show production-safe patterns.

5. Get the page-level basics right

A unique, specific title that says what the page is. Google writes the title links shown in results automatically, but your <title> is a primary input.
A meta description that summarizes the page. It can be used as the snippet people see before they click. Google said in 2007 that descriptions don't affect ranking.
One clear main heading that matches the page's purpose.
A self-referencing canonical tag on the production domain, never pointing at staging.
Crawlable links. Google says it can generally only crawl a link if it's an <a> element with an href attribute. Buttons that navigate with JavaScript alone may never be followed.
Main content in the HTML. Some site builders render pages in the browser with JavaScript. Make sure your text and links are in the HTML that Google receives. Phase 2 shows how to check.
The same content on mobile. Google uses the mobile version of a site for indexing, so text, links, and structured data hidden on phones are missing for Google too.
Real images with descriptive alt text for anything meaningful. Google doesn't index CSS background images.
One preferred address. Choose HTTPS and either www or the bare domain. Redirect every other variant to it in one hop.
A real 404. Missing pages should return a 404 status, not a "page not found" message served with a 200 status, and not a redirect to your homepage.
If you serve a local area, create and verify a Google Business Profile, with the same business name, address, and phone number as your site.

Skip the old rituals. Google Search doesn't use the keywords meta tag, and the same guide says "the length of the content alone doesn't matter for ranking purposes." Write as much as the topic needs and no more.

6. Ship a clean XML sitemap and structured data that matches the page

List only canonical URLs that return a 200 status and are meant to be indexed in your sitemap: no redirects, no noindex pages, no staging hosts. Add a Sitemap: line to robots.txt. Google's sitemap documentation names "your site is new and has few external links to it" as a situation where a sitemap helps. It is still a hint, not a guarantee of crawling or indexing.

Organization markup with your logo and sameAs links to your real profiles helps Google connect your brand's identity. Google's rules say: "Don't mark up content that is not visible to readers of the page." Don't add markup in the hope of a visual result. FAQ rich results are a good example: Google limited them to government and health sites in 2023, and its FAQ structured data documentation says they stopped appearing in Search on May 7, 2026.

7. Decide who can read the site, including AI search

Google's AI Overviews and AI Mode use the same index as regular results. Google says "there are no additional requirements" to appear in them: a page must be indexed and eligible to show a snippet. So avoid nosnippet on pages you want quoted.

Other assistants run their own crawlers. OpenAI, for example, says that sites that opt out of its OAI-SearchBot won't be shown in ChatGPT search answers, though they can still appear as navigational links. Decide which crawlers you allow in robots.txt, then check your CDN or firewall settings, because some block AI crawlers by default regardless of robots.txt. Our guide to whether AI assistants can read your website walks through the checks, and PolyDraft's AI visibility overview explains how we approach it.

If the domain was used before, check its past too: what it hosted (the Internet Archive's Wayback Machine), which sites link to it, and, once Search Console is verified, the Manual actions and Security issues reports.

Phase 2: the launch-day SEO checklist, proved on production

Staging checks don't count. Run these against the live domain the hour you launch, and write down the result of each one.

Check	How to verify	Pass condition
Pages load	Open the homepage and three key pages; check the HTTP status	200, correct content, no redirect loop
Indexing allowed	View source and response headers on each key page	No `noindex` in the robots meta tag or `X-Robots-Tag` header
Content in HTML	View page source, then URL Inspection's rendered HTML after Search Console is set up	Title, main text, and internal links are present
Canonicals	View source	Self-referencing absolute URLs on the production domain
robots.txt	Open `/robots.txt`	No site-wide `Disallow`; a `Sitemap:` line pointing to the live sitemap; the crawlers you want are allowed
Sitemap	Open it and request every URL	Production URLs only, each returning 200
One address	Try `http://`, `www`, and the trailing-slash variant	One hop to the preferred HTTPS URL
Missing pages	Visit a made-up path in every section, such as `/blog/`	Returns a 404 status, not a redirect
Mobile view	Open key pages on a phone	Same main content, navigation, and links as desktop
Speed	Run PageSpeed Insights on each page template	No blocking issues in the lab results. A new site has no Core Web Vitals field data at first, and missing data isn't a pass
Search Console	Verify a domain property, submit the sitemap, run URL Inspection's live test on key pages	Sitemap accepted; live test says the page can be indexed
Bing Webmaster Tools	Import the site from Search Console and submit the sitemap	Sitemap processed
Analytics	From a phone with no ad blocker, accept the consent banner and complete your main conversion once	The visit and the conversion each appear once, in the production property
Internal links	Crawl the site from the homepage	No broken links; every key page reachable

Write down the launch-day numbers, even when they are zero. That dated baseline is what you compare against in month two.

What it showed when we ran it on polydraft.ai on October 13, 2026:

Check	Result
Pages load	Pass: homepage and five key pages returned 200
Indexing allowed	Pass: robots meta `index, follow`, no `X-Robots-Tag` header
Content in HTML	Pass: titles, body text, and links are in the server-rendered HTML
Canonicals	Pass: each of those six pages points to itself on the production domain
robots.txt	Pass: all crawlers allowed, with named rules for seven AI crawlers, and a `Sitemap:` line. We didn't check CDN bot settings
Sitemap	Pass: all 55 URLs returned 200 (one returned a 502 on the first request and 200 on a retry)
One address	Partial: trailing-slash URLs redirect in one hop. We didn't test `http://` and `www` in this run
Missing pages	Fail: a made-up URL at the root returned 404, but made-up URLs under `/blog/` and `/compare/` redirected to the section page. That is a soft-404 pattern. The fix is logged for our site template
Internal links	Pass with one exception: the 55 sitemap pages link to 59 internal URLs, and one pointed to a post that wasn't published yet, so it redirected to the blog index until that post went live
Mobile view, Speed, Bing Webmaster Tools	Not part of this run
Search Console, analytics	Not part of this run; they need account access

The point of writing results down is to find failures like the redirect above. It looked fine in a browser and only showed up when we requested a URL that shouldn't exist.

Phase 3: weeks one to four, watch, fix, earn

How a new URL moves through Google Search: discovered, crawled, indexed, then shown for queries, with the two most common stuck states explained.

Know where each page is stuck

Every URL moves through the same stages: Google discovers it, crawls it, indexes it, and then shows it for real queries. Google's recrawl documentation says crawling "can take anywhere from a few days to a few weeks." Check weekly in Search Console's Page indexing report, and use URL Inspection for individual pages.

If a page isn't on Google, diagnose by status:

What Search Console shows	Likely cause	What to do
Excluded by `noindex`	A tag or header left over from staging	Remove it, then request indexing once
Blocked by robots.txt	A `Disallow` rule covers the URL	Fix the rule; remember robots.txt doesn't remove indexed URLs
Alternate page with proper canonical	The canonical points to another URL	Make the canonical self-referencing if this page should rank
Discovered – currently not indexed	Google found the URL but rescheduled the crawl	Check Settings > Crawl stats for server errors first, then link to the page from indexed pages and keep it in the sitemap
Crawled – currently not indexed	Google fetched it and chose not to index it for now	Make it clearly more useful than similar pages, or merge it. Resubmitting won't change the outcome

Request indexing once for your most important pages after launch. Repeated requests don't make crawling faster. Google's Indexing API is only for job posting and livestream pages, so services that promise instant indexing through it are a risk, not a shortcut.

Improve before you add

Our rule of thumb for the first month: improve the pages you have before publishing more. A few strong pages get crawled, linked, and judged as a set more easily than many new ones. Read the queries that start appearing in the Performance report. If people find a page through questions it barely answers, answer them on that page. Add a new page only when it serves a distinct need better than anything on your site.

Earn your first real mentions

New sites rarely get links by accident. These are legitimate first sources, roughly in the order they pay off:

Profiles you control: company pages on LinkedIn, GitHub, Crunchbase, or the networks your customers use. List them as sameAs links in your Organization markup.
Partner and integration directories where you genuinely belong.
Curated category directories that real buyers use, not link farms.
Launch platforms and communities where you already take part.
Founder talks, podcasts, and guest expertise.
Customers' "tools we use" pages, if they'd mention you anyway.

Don't buy links or trade them at scale. Google's spam policies treat link schemes as manipulation.

Log every change with a date

Keep a simple log: date, URL, what changed, why. Search data arrives with a delay and many things move at once. A dated log is the most reliable way to look back later and form a reasonable hypothesis about what helped. Even then, treat a before-and-after chart as a clue, not proof.

How long until a new site ranks? Three clocks

"How long" mixes three questions with different answers:

Indexing: days to weeks. Google's recrawl documentation gives that range for crawling, and URL Inspection shows when a specific page is on Google.
First impressions: usually weeks. The first queries are often your brand name and narrow long-tail searches. Treat this as a practitioner estimate: we'll replace it with polydraft.ai's own Search Console dates once we have them.
Top-10 rankings for non-brand searches: months, if at all. In the original 2017 sample of Ahrefs' study on how long pages take to rank, 5.7% of new pages reached the top 10 within a year. Treat any such figure as a base rate, not a forecast for your site.

Google hasn't confirmed a "sandbox", but new sites do start without the signals older sites have built up, and rankings can move around while those signals form. The practical answer is the same: there's no fixed waiting period to sit out.

When	What to check	Where
Launch day	Every row of the launch checklist passes	Browser, crawler, Search Console
Weeks 1–2	Key pages crawled and indexed	URL Inspection, Page indexing report
Weeks 3–4	First impressions, often for your brand name and narrow queries	Performance report
Months 2–3	Which pages earn impressions and clicks, and whether visitors convert	Performance report, analytics
Months 4–12	Whether positions stabilize for your target queries, and which pages need improving or merging	Performance report, compared month over month

These are the order in which signals become checkable, not a schedule Google follows. Some sites move faster, many move slower, and nobody can promise dates.

Mistakes that waste a new site's first months

Launching with a staging noindex or a Disallow: / in robots.txt
A sitemap full of redirects, noindex pages, or staging URLs
Missing pages that redirect to a section page or the homepage instead of returning 404
Publishing dozens of near-identical pages to "cover keywords"
Targeting broad head terms your new domain can't compete for yet
Changing URLs a few weeks after launch, which forces redirects and makes Google re-crawl and reassess the moved pages
Chasing word counts, keyword density, or meta keywords

Where PolyDraft fits

The PolyDraft site scaffold includes canonical tags, an XML sitemap, robots.txt, Open Graph tags, and structured data support, which cover several Phase 1 items. They are starting points, not a substitute for the launch-day checks, as our own soft-404 finding shows. Run the checklist on your rendered production pages whichever builder you use, and verify Search Console in your own Google account. If you're still choosing a tool, our guide on how to evaluate an AI website builder lists what to test before you commit. Early-stage teams can see how we approach this on PolyDraft for startups.

Frequently asked questions

How long does it take for a new website to show up on Google?

Indexing usually takes days to weeks, and Google doesn't guarantee that any page will be indexed. Ranking in the top 10 for non-brand searches takes months, and many new pages never get there. Check individual pages with URL Inspection rather than searching for them.

Is there a "Google sandbox" for new websites?

Google has never confirmed one. New sites start with few links, little history, and few signals that their pages are the best answer, and rankings can fluctuate while those build. There's no fixed penalty period to wait out.

Does a small new website need a sitemap?

It isn't required. Google lists new sites with few external links as a case where a sitemap helps, and it costs little to add. Keep it accurate.

Does a new website need anything special for AI search?

Not for Google: AI Overviews and AI Mode use the regular index, so a page has to be indexed and allowed to show a snippet. For other assistants, allow the crawlers you want in robots.txt and check that your CDN doesn't block them.

How many pages should a new website have at launch?

As many as there are distinct needs you can answer well. For most startups that means a homepage, one page per core offer, prices, about, and contact. Add pages when real demand appears.

Should I request indexing for every new page?

No. Submit your sitemap and request indexing for a handful of key pages. Repeated requests for the same URL don't speed up crawling, and Google's Indexing API only covers job posting and livestream pages.

Continue with the next task

After launch, the work changes from setup to an ongoing loop. Our guide to post-launch SEO covers that phase.
Planning which pages to build: Website architecture for SEO
Launching with a single page: One-page website SEO
Using AI to help with the work, safely: How to use AI agents in SEO

About this article. Drafted with AI assistance from PolyDraft's internal research notes, checked against the sources below, and reviewed and approved by Vincent Chen. The polydraft.ai figures come from Ahrefs data recorded on August 29 and October 5, 2026, our Ahrefs keyword research of October 5, 2026, and the launch checklist we ran on polydraft.ai on October 13, 2026. We don't claim any ranking or traffic results. The diagrams are original illustrations made for this article, not screenshots of customer sites.

Sources

Accessed October 2026.

Google Search Central: How Search works
Google Search Central: Creating helpful, reliable, people-first content
Google Search Central: Spam policies for Google web search
Google Search Central: SEO Starter Guide
Google Search Central: Influencing your title links and Control your snippets in search results
Google Search Central Blog: Improve snippets with a meta description makeover (2007)
Google Search Central: Mobile-first indexing best practices
Google Search Central: Google Images SEO best practices
Google Search Central: Introduction to robots.txt and Block indexing with noindex
Google Search Central: Make your links crawlable
Google Search Central: Sitemaps overview
Google Search Central: Organization structured data, General structured data guidelines, and FAQ structured data
Google Search Central: AI features and your website
Google Search Central: Ask Google to recrawl your URLs and Indexing API quickstart
Search Console Help: Page indexing report, URL Inspection tool, and Crawl stats report
Google Business Profile Help: Tips to improve your local ranking on Google
OpenAI: Overview of OpenAI crawlers
Ahrefs: How long does it take to rank in Google?
PageSpeed Insights: pagespeed.web.dev

SEO for a New Website: A Zero-to-Indexed Launch Plan