Check this page with an assistantOpens a chat asking it to summarise this article and name the evidence behind each claim.
Claude opens with the prompt on your clipboard: Anthropic does not support prefilled prompts on the web, and we would rather copy it than ship a button that drops it.
Your First Problem Is Discovery, Not Ranking
A new site with no impressions is almost never being outranked. It is usually not in the running, because Google has not found the pages. These are entirely different problems with entirely different fixes, and confusing them costs new sites their first quarter.
The test takes thirty seconds. Inspect an interior URL in Search Console and read the coverage state. "URL is unknown to Google" means the crawler has never seen it, which is a discovery failure. "Crawled, currently not indexed" means it was seen and not selected, which is a quality or duplication question. "Submitted and indexed" with no impressions means you are in the index and losing, which is the only one of the three that is a ranking problem.
This distinction matters more than any content advice on a new site, because the failure is silent. Nothing warns you. You publish, the site looks finished, analytics shows nothing, and the natural conclusion is that the content needs to be better or that SEO takes time. Meanwhile the actual state is that one URL is known and the other ninety-nine have never been requested.
You have inspected at least three interior URLs and know which of the three states your site is actually in.
Week One: Make the Site Findable
Five actions, in order, none of which involve writing anything. Everything in the following twelve weeks depends on these working, and they take an afternoon combined.
- Verify the property in Search Console.
Prefer a domain property, which covers every subdomain and both protocols, so you are not later confused by data split across variants. This is also the only way to see the coverage states described above.
- Submit a sitemap.
Generating a sitemap and submitting it are two different things, and the gap between them is where a surprising number of sites live. A sitemap that exists at a URL nobody has declared is a file, not a signal. Submit it explicitly and confirm the report shows it read.
- Check robots.txt is not blocking you.
Staging configurations survive launches with depressing regularity. A blanket disallow, or a stray noindex header left on a template, will cheerfully undo everything else on this list without any visible symptom on the site itself.
- Link deeply from the homepage.
Crawlers follow links. If your homepage links only to three marketing pages and the rest of the site is reachable through a search box or a client-side filter, discovery stops there. Real anchor-linked routes to every section are the requirement, not a navigation aesthetic.
- Get one or two real external links.
A single link from any indexed page gives crawlers an independent route in. Your own profiles, a supplier listing, a directory you legitimately belong to. This is discovery infrastructure rather than link building, and the easy sources are listed in how to get backlinks.
Verifying That It Worked
Do not assume submission equals discovery. A week after submitting, inspect three URLs from different sections and confirm each is known. If they are still unknown, something in the chain is broken and more content will not fix it.
| Coverage state | What it means | Next action |
|---|---|---|
| URL is unknown to Google | Never discovered | Sitemap submission, internal links, one external link |
| Discovered, currently not indexed | Known but not yet crawled | Usually wait; check for crawl blocks and server errors |
| Crawled, currently not indexed | Seen and not selected | Check duplication and thinness; see duplicate content |
| Submitted and indexed, no impressions | In the index, not competitive | Now it is a search intent and relevance question |
| Excluded by noindex or robots | You told Google not to | Fix the directive; see robots vs noindex vs canonical |
Expect partial coverage rather than all-or-nothing. Google indexes what it considers worth indexing, and a new site with a hundred pages may see twenty picked up first. That pattern is normal. What is not normal is a month passing with a submitted sitemap and no interior URL known at all, which points at a technical block rather than patience.
The Sandbox Question, Answered Honestly
Google has consistently denied running a sandbox that deliberately holds new domains back. Practitioners consistently observe that new domains take months to rank stably on anything competitive. Both of these can be true, and the disagreement is about mechanism rather than about what you will experience.
The plausible non-conspiratorial explanation is that ranking systems need evidence, and a new domain has none. No history of satisfying queries, no accumulated references, no track record. A system with no evidence about you will be conservative about putting you in front of people, and that conservatism looks identical from the outside to a deliberate filter. Whichever it is, the remedy is the same: accumulate evidence.
Published observations cluster around three to six months before stable movement on competitive terms, and six to twelve before search contributes meaningfully to a business. Treat those as the shape of the curve rather than as a schedule, since none of them come from disclosed datasets. The important planning consequence is that a new site should not be evaluated on month-three rankings, and anyone selling you a faster outcome is selling something that has repeatedly not existed.
Weeks Three to Twelve: Give the System Something to Assess
Once discovery works, the job changes to producing evidence. Specific pages answering specific questions, a clear identity that resolves consistently, and enough structure that new pages get found automatically rather than one at a time.
Start with the queries where you can genuinely be the best result. That means your own name, your specific service in your specific place, and narrowly qualified questions where the current results are weak. These are not consolation targets: they are the only queries a new domain can win quickly, and winning them produces the impressions data that makes every later decision evidential rather than speculative. The selection method is in long-tail keywords.
At the same time, make your identity unambiguous. One business name, one description, one address, spelled the same everywhere it appears on and off the site. New domains suffer disproportionately from identity ambiguity because there is no history to resolve it against, which is the practical argument for the work in entity SEO. And decide your URL structure now, while changing it is free, using site architecture.
What Cannot Be Hurried, and What People Sell Anyway
Accumulated evidence takes time by definition. No amount of publishing compresses it, and the tactics marketed as accelerants are mostly the ones that create long-term problems: bought links, expired domains with a history you did not earn, and mass-generated pages.
The specific temptation for new sites is volume, on the theory that more pages means more chances. On a domain with no track record it tends to produce the opposite: a large number of pages that nothing distinguishes, published quickly, on a site with no history. That is a recognisable pattern and not a flattering one, and it is precisely what the scaled-content provisions in Google's spam policies describe. Twelve pages that each answer something properly is a better first quarter than a hundred that do not.
What does compound is the boring half: pages that stay accurate, identity that stays consistent, and a structure that keeps working as you add to it. Set the reporting expectation accordingly before anyone starts asking why traffic is flat, using the framing in how long SEO takes for a small business and the reporting framework.
Questions People Ask About New Websites
- How long does it take Google to index a new website?
Initial pages are commonly indexed within a few days to a few weeks once Google knows they exist, with fuller coverage of a site taking a couple of months. The variable that matters most is not time but discovery: a page Google has never been told about can sit unindexed indefinitely, which is why submitting a sitemap is the highest-value action on a new site.
- Does the Google sandbox exist?
Google has consistently denied operating a sandbox that deliberately suppresses new domains. What practitioners observe is real enough: new sites commonly take three to six months before ranking stably on competitive terms. Whether that is a filter or simply the time required to accumulate the signals any assessment needs is unresolved, and the practical advice is the same either way.
- What should I do first on a brand new website?
Verify the property in Search Console and submit a sitemap. Everything else is downstream of Google knowing your URLs exist. A site can have a hundred excellent pages and receive zero impressions indefinitely if nothing has ever told a crawler about them, and this is a far more common failure than people expect.
- Why is only my homepage indexed?
Because that is the only URL discovery has reached. Google finds pages by following links and by reading sitemaps, so if your homepage does not link deeply into the site and no sitemap has been submitted, the crawler has nowhere to go. Check by inspecting an interior URL: 'URL is unknown to Google' means discovery, not quality, is the problem.
- Should I target competitive keywords on a new site?
Not first. Target the specific, low-contest queries where your page can genuinely be the best answer, including your own brand name and narrowly qualified questions. Competitive head terms on a new domain consume months and usually produce nothing, while specific queries can begin returning traffic within weeks and give you data to work from.
