A new website becomes eligible for Google Search when Google can discover its important URLs, fetch and understand their content, and decide they are worth indexing. You can make each stage easier: verify the public site in Search Console, publish a consistent sitemap, expose crawlable navigation, remove accidental indexing barriers and provide pages that solve real visitor questions. None of these five steps guarantees an indexing date or a rank. Google's sitemap guidance explicitly distinguishes discovery from guaranteed crawling and indexing.
Step 1: establish the public version and verify Search Console
Before submitting anything, decide which site visitors should use. Does http:// redirect to HTTPS? Does the preferred hostname include www? Is the trailing-slash policy consistent? Does a visitor who clicks the homepage reach the actual published site rather than a staging hostname? A new site with four public variants and conflicting canonicals gives crawlers extra decisions to make; a URL-prefix Search Console property also covers only its precise prefix, while a verified domain property covers variants. Check Google's property-type guidance before interpreting an empty report as evidence that Google knows nothing about the site.
Verify the property you can manage. Record the final URLs of a handful of representative pages: homepage, core service or product, category and first useful article. Inspect each exact URL in Search Console. The initial result shows Google's indexed knowledge, which may be old; Test live URL fetches a current version but cannot predict canonical selection or quality-based indexing. Google: URL Inspection tool.
Search Console will not create a site or publish missing content. If the homepage resolves but its key product pages are unpublished, publish and link them first. Do not treat a verified property badge as confirmation that the sitemap is readable, that every page is indexed or that your pages rank.
Step 2: submit a sitemap of the pages you actually want found
Find your existing sitemap before creating a second one. Many CMSs produce one automatically; a hand-written file may go stale as soon as URLs change. Include preferred, canonical URLs that return useful public content. Exclude staging pages, password-protected previews, search-result parameters, deleted pages and URLs deliberately marked noindex. Google describes a sitemap as a way to tell it about important URLs, not a compulsory ticket or an indexing guarantee. New sites with few external links can especially benefit because discovery via other sites may be limited. Google: Learn about sitemaps.
Open the XML in a browser, check its HTTP status and read several <loc> entries as actual URLs. Does each resolve to the same final page used in navigation? Is the declared canonical compatible with it? Submit the file through Search Console's Sitemaps report and look for fetch/parse errors. A sitemap index is appropriate when the site truly has multiple child sitemaps, not an invented “priority channel.” If you use <lastmod>, update it for significant content changes; Google ignores <priority> and <changefreq>. Google: Build and submit a sitemap.
For a few particularly important new URLs, URL Inspection allows a manual indexing request after the page is actually accessible. For many URLs, the sitemap is the appropriate submission route. Repeated requests for the same unchanged URL will not accelerate a crawl; Google does not promise that requesting a recrawl results in index inclusion. Google: Ask Google to recrawl your URLs.
Step 3: make the site navigable and technically readable
Offer normal paths to important content
A sitemap is not a substitute for navigation. From the homepage, can a person reach the primary services, products or topics? Does a category or topic hub link to its articles? Are links normal <a href> elements rather than click handlers on non-link boxes? Can the main text, headings and links be seen on a mobile device without authentication or a mandatory interaction? Google documents how crawlable anchors help it discover linked pages. Google: Make links crawlable.
Keep URLs human-readable where practical, but do not rename a stable route merely because it includes a date or a harmless identifier. If you must move a live page, redirect the old address to a genuinely corresponding new page and update links and sitemap. Redirecting every discarded URL to the homepage is not a valid substitute for a missing page. Google treats permanent redirects as a strong signal for the target canonical, whereas a 200 response containing an error-looking page can be considered a soft 404. Google: redirects; Google: HTTP status codes.
Remove launch-stage blocks in the right order
Examine production /robots.txt and the exact URL's meta robots and HTTP X-Robots-Tag. A leftover Disallow: / can prevent crawling. A noindex on a publicly crawlable page prevents search indexing. If you both disallow crawling and add noindex, Googlebot may never see noindex; robots.txt is not a reliable method of keeping a URL out of search results. Google: noindex guidance.
Use the final public response as evidence. An ordinary content page returning 200 allows processing but does not guarantee inclusion; 404/410 means gone, a 403 may indicate a permission error, while sustained 5xx and 429 responses signal server trouble and can slow crawling. Verify the mobile rendered body and resources in the live inspection when possible; a blank shell that fills only for logged-in users is not a useful public page. Google: How HTTP status codes affect crawlers.
Keep canonical signals aligned
Set a coherent preferred URL for duplicates such as HTTP/HTTPS, tracking-parameter or www variants. A self-referential canonical on a preferred page is useful, but do not canonicalize every article to the homepage. Align sitemap entries and internal links to your chosen public URLs; canonical tags are strong preferences, not guaranteed commands. Search Console's Google-selected canonical can differ from your declared value. Google: canonical methods and limitations.
Multilingual sites need another check: a Chinese and an English article may target corresponding topics but each language should ordinarily have its own indexable canonical, with correctly paired localized-version signals where used. Do not use hreflang to repair a blocked or redirected English URL. Google's localized-version guidance explains how alternate-language URLs should reference each other.
Step 4: publish pages that stand on their own
A new site's first pages should explain the business and the questions its intended visitors genuinely ask. A service page might need scope, suitability, constraints, process and an actionable contact route. An article might need prerequisites, a procedure, conditions, an example and what to do when the procedure fails. There is no Google rule that a site must start with exactly five or ten posts, and a large batch of paraphrased guides is not a substitute for a coherent website. Google asks whether content adds original information, describes the topic substantially and helps readers meet their goals; it does not prescribe a word count. Google: Creating helpful, reliable, people-first content.
For example, a new B2B website selling a technical component can publish one complete specification and compatibility page, one purchasing/lead-time page grounded in actual policy, and a troubleshooting guide for a real application. This is a planning illustration, not a claim about a client site or a formula that triggers Google's crawl. Make the claims that your company can verify. Do not create synthetic screenshots of Search Console, fabricated ratings or “indexed in seven days” testimonials to make the page appear authoritative.
Research the questions people actually ask, then map one primary intent to each page. Compare representative search results in the target language and market: do users expect instructions, vendor listings, definitions or a comparison? A page titled “cost” that never states scope, pricing model or required inputs fails that intent even if it repeats “cost” many times. Related questions can be covered within one strong guide rather than one thin URL per synonym. Use a relevant image only when it conveys something accurate; image discovery needs more than alt text alone. Do not decorate a page with fake analytics graphics.
Step 5: monitor the funnel from discovery to clicks
Inspect important URLs and the Page indexing report. “Unknown” suggests discovery work; “Discovered – currently not indexed” means known but not crawled; “Crawled – currently not indexed” needs a look at the fetched page, duplicate selection and usefulness. If Google selected a different canonical, inspect that page instead of repeatedly resubmitting this one. A successful live inspection does not override a different indexed canonical or force an inclusion decision. Google: URL Inspection states.
Once pages are indexed, use Performance to examine queries, landing pages, country and device. Few impressions suggest query relevance and search competition may need work; impressions without clicks suggest result appearance, position or intent mismatch. Search Console and analytics measure different stages: analytics sessions cannot reliably establish what URL Google selected as canonical. Read the Search Console Performance guide for its filters before applying one CTR target to every page.
| Observation | A sensible first response | What it does not prove |
|---|---|---|
| Sitemap accepted; URL not indexed | Inspect the URL, links, fetched page and canonical. | That the sitemap failed or a penalty exists. |
| Live test says URL is available | Compare the current page with the indexed result. | That it will be indexed or rank. |
| Many 5xx in crawl diagnostics | Stabilize the server and review logs. | That another submission fixes availability. |
| Indexed but no target-query impressions | Review query intent and the depth of the page. | That the page was not discovered. |
| Impressions but few clicks | Inspect the real SERP and title promise. | That an indexing setting is broken. |
Common shortcuts that do not deliver the promised outcome
Do not set the Indexing API up for ordinary articles, products or a company homepage. Google restricts its Indexing API to pages containing JobPosting or BroadcastEvent embedded in VideoObject; a successful API response merely acknowledges notification, not inclusion in the index. Google: Using the Indexing API.
Do not promise “index within 72 hours” from a sitemap, daily checks or paid links. A link from a genuinely relevant site can give people and crawlers a route to your work, but a fabricated package of “high-indexing-rate” links is not evidence of a repeatable outcome. Share useful resources where they fit and assess link partnerships for relevance and transparency, not made-up percentages or an alleged universal anchor-text ratio.
Do not treat “crawl budget” as the default explanation for a small new site. First solve actual server errors, disconnected pages, duplicate URLs and low-value content. Google says crawl-budget optimization is primarily relevant for very large or fast-changing sites; most smaller sites should focus on making important content accessible. Google: optimize crawling performance.
For a deeper discussion of slow indexing, no impressions and no clicks as separate problems, continue with our new-site SEO troubleshooting guide. If you need help shaping a site's pages and technical foundations into a coherent search program, the Google SEO service overview describes the business service. The useful milestone is not an arbitrary launch-week percentage; it is an accessible site with distinctive pages and a documented way to find and fix each real obstacle.