New Site SEO 2026: Rank a Brand-New Domain

Administrator in Search Engine Optimization Sep 08 / 1 week ago

Here is the number that quietly ruins most launch plans. Ahrefs studied one million URLs its crawler first saw in September 2023 and found that only 1.74% of them reached Google's top ten inside a year. Not the top three. The top ten. Of the pages that did break through, the largest single group arrived inside the first month, and the rest trickled in across the eleven that followed.

That stat is not a reason to quit. It is a reason to plan differently. A brand-new domain has no crawl history, no links, and no record of a single person choosing it in a search result. Google has almost nothing to judge. So it does the sensible thing and waits for evidence.

This guide covers new site SEO across those first 90 days. Not theory. A week-by-week plan, a straight answer on the sandbox debate, the checks to run before you buy a domain, and the part almost nobody explains: how to read progress while your click count still says zero.

What this 90-day plan actually promises

By day 90 a healthy new domain should be fully indexed, crawled on a steady rhythm, ranking somewhere for a handful of long-tail queries, and feeding you clean data. Clicks may still be tiny. That is normal. Small traffic at day 90 is not failure.

Four opinions shape the rest of this piece, and I will defend each one:

• The Google sandbox is not a filter. It is the absence of data plus a very small crawl budget, and those two things behave almost identically from the outside.
• "How long does SEO take" has no single answer. It has a distribution, and quoting a flat number is dishonest.
• Buying an aged domain moves your risk around rather than removing it.
• On a fresh domain, crawl behaviour and coverage state tell you more than clicks do. Good new site SEO in the first quarter is measured in crawl signals, not revenue.

The plan assumes a small team, a modest budget, and no existing brand. If you are launching under a known company name, everything below happens faster, because brand search does a lot of the trust work for you.

Is the Google sandbox real, or just missing data?

No switch inside Google holds new domains back for six months. What exists is arguably worse. A fresh domain has no links, no click history, and a very small crawl allocation. The delay is completely real. The punishment is imaginary.

John Mueller has addressed this in Google Search Central office hours more than once. His position is that Google runs systems that try to work out where a new site fits alongside everything already known, and from the outside that process looks like a filter. I think the sandbox label survives because it is emotionally easier. Blaming a secret gate feels better than admitting a site has not yet earned a reason to rank.

There is a clean test for this. If a true sandbox existed, nothing new could rank quickly. Yet breaking news pages, Reddit threads, and Product Hunt launches rank within hours. Ahrefs found that 40.82% of the pages that eventually reach the top ten arrive there within a month. Speed is clearly possible. It just needs signals a blank domain does not have on day one.

How long does new site SEO really take?

The truthful answer is a range, not a date. Most new pages never reach page one at all. Among those that do, roughly four in ten arrive within the first month, and the remainder are spread thinly across the rest of the year. Anyone promising a flat "three months" is selling certainty they do not own.

Treat each published page as a ticket in a lottery with poor base odds and one useful property. You can buy more tickets, and you can make each ticket better. Ten strong pages beat one hundred weak ones, because the weak ones drain crawl budget and teach Google that your site is mostly filler.

Two variables move the odds hard. First, competition. A page about payroll software in Germany faces different resistance than a page about a niche hardware fault code. Second, existing trust. A funded startup with press coverage gets discovered faster than a side project with no mentions anywhere. The full Ahrefs data set is worth reading in the original: Ahrefs on how long it takes to rank and how old top pages are.

What should you actually do in week one?

Week one is plumbing, not content. Your only goal is to make the site easy to crawl, easy to verify, and easy to measure later. Skip this and every later diagnosis becomes guesswork, because you will have no baseline to compare against.

• Verify the domain in Google Search Console and in Bing Webmaster Tools. Both are free and both take minutes.
• Publish a real home page, an about page naming actual humans, and a working contact page.
• Submit one sitemap and confirm Search Console reads it without errors.
• Check that DNS resolves cleanly and the TLS certificate is valid on every host name.
• Choose one canonical host, with www or without, and redirect the other.
• Install Google Analytics 4 and write your launch date down somewhere permanent.

Two build habits pay off later. Keep your launch pages light, because images are usually the single largest contributor to page weight, and our guide to picking the right image format, sizing, and alt text covers the essentials in about an hour. Then use proper heading order and real link text from day one, since retrofitting is miserable. Our guide to WCAG 2.2 and the European Accessibility Act, which is upfront that accessibility is not a direct ranking factor separates what EU law requires from what genuinely helps search. Solid new site SEO starts with a site that is boringly correct.

A word on tools, honestly. Google Search Console is non-negotiable and free, though its data lags roughly two to three days. Bing Webmaster Tools is thinner but worth the ten minutes, because Bing's index feeds several AI assistants. Screaming Frog crawls 500 URLs free, which covers most new sites, and the interface looks dated but runs fast. Ahrefs and Semrush cost more than most launch budgets expect. Ahrefs Lite runs about 129 USD a month and Semrush starts around 139 USD, which is genuinely hard to justify in month one when you have no rankings to track yet. Ahrefs does sell a cut-down Starter tier near 29 USD if all you want is keyword data.

Google's own guidance is refreshingly calm about all this. The Google Search Central SEO starter guide states plainly that most sites are found and added automatically, and that some changes take hours while others take months.

What matters most in weeks two to four?

Now you publish. Aim for eight to twelve genuinely useful pages, each aimed at a specific long-tail question with visible search demand. Resist the urge to chase your head term. You will lose that fight for months, and losing it teaches you nothing.

Pick queries where the current page one is weak. Thin affiliate posts, outdated 2022 guides, and forum threads are all signs of a gap. Before you write, work out exactly what the searcher wants, because a mismatch here wastes the whole page. Our breakdown of how to match the precise search intent behind a query covers the micro-intent layer that most briefs miss.

Length is a trap in this window. Write until the question is fully answered, then stop. If you need convincing, read why word count is not a ranking factor and what to measure instead. Before each page goes live, run it through the on-page SEO checklist you can apply to every page before publishing. Ten minutes per page here prevents a month of avoidable fixes.

What changes in weeks five to eight?

This is when the site stops being invisible. Crawl frequency settles into a pattern, impressions start appearing in Search Console, and your first rankings show up between position 20 and 60. Those positions look useless. They are actually your first real feedback.

Your job now is depth and connection. Cluster your pages so related articles link to each other and up to a hub page. That structure is the fastest way to signal what your site is about, and our guide to building topical authority with a pillar page and focused cluster pages lays out the hub-and-spoke pattern.

Technical quality starts to matter more once pages compete. Check your current Core Web Vitals thresholds, including INP and LCP, on real devices rather than a fast laptop. And remember that the mobile version of your page is the one Google indexes, so any content hidden on small screens is content Google may never see. Good new site SEO treats the phone view as the source of truth.

What does month three look like?

Month three is triage. You now have eight weeks of data showing which pages got crawled, which earned impressions, and which sat untouched. Kill nothing yet. Improve the near-misses, expand the winners, and diagnose the pages Google ignored entirely.

This is also the first sensible moment to think about links and mentions. Not before. A brand-new domain buying links in week two looks exactly like every spam site that ever existed. Real mentions come from doing something worth mentioning: original data, a free tool, a genuinely better explanation. When you are ready, our review of link building tools worth the subscription cost compares the realistic options.

Two other levers matter more in 2026 than they did three years ago. Author credibility, covered in our guide to the trust signals behind E-E-A-T, and visibility inside AI answers, covered in our practical guide to getting surfaced in AI generated answers. New sites often get cited by ChatGPT, Perplexity, or Gemini before they rank well in classic search, because those systems weight clear answers heavily. That shift also changes what a ranking is worth, which our guide to winning visibility when AI Overviews answer the query on the results page works through in detail.

Should you buy an expired or aged domain?

Usually no. An aged domain can carry real link equity, but it can also carry spam history, a manual action, or a reputation attached to a completely different topic. You are buying an unknown past, and the discount is rarely worth the uncertainty.

Run three checks before any purchase, in this order. First, read the domain's history in the Wayback Machine from the Internet Archive. A former casino affiliate site registered in Malta is a poor base for a fintech brand in Berlin, and the archive shows you that in two minutes. Second, pull the backlink profile in Ahrefs, Moz, or Semrush and judge whether the links are real editorial mentions or an obvious farm. Third, confirm the domain is not flagged anywhere, using our guide on how to check whether a domain sits on an email or safe-browsing blacklist and how to get removed.

Weigh that against the alternative. A clean registration costs roughly 10 to 15 USD, or about 9 to 14 EUR, a year and carries no baggage at all. Cloudflare Registrar sells .com at cost, near 10 USD with no renewal markup. Namecheap sits around 10 USD in year one and closer to 15 USD on renewal. GoDaddy discounts the first year then renews above 20 USD, so read the renewal price rather than the headline. Aged domains routinely sell for 500 USD and up. My view is that unless the archive and the link profile are both genuinely clean and genuinely relevant to your topic, the fresh domain wins on risk-adjusted value every time. Starting clean also keeps your new site SEO diagnosis simple, because every signal you later read belongs to you.

What should a brand-new site never do in month one?

Most launch damage is self-inflicted. Five mistakes account for the bulk of it, and each one is tempting precisely because it feels like fast progress. Avoiding them costs nothing, which makes them the cheapest wins available to you.

• Mass link buying. A domain with no traffic and 400 new links in three weeks is a pattern, not a strategy. Wait until you have something worth linking to.
• Programmatic page dumps. Generating 5,000 near-identical location or comparison pages on a domain with no trust burns your crawl budget and signals low quality.
• Chasing head terms. Writing "best CRM software" in week two is a guaranteed loss. Ahrefs found the average number one result is about five years old, and roughly three quarters of top ten pages are older than three years.
• Constant redesigns. Changing URLs three times in 90 days resets discovery each time. Pick a structure and keep it.
• Chasing novelty files. Our evidence-based look at llms.txt, which concludes it is optional and low-impact for most sites right now is the clearest example. Fix crawling and content first, before adopting proposals no major AI provider has committed to reading.

There is a sixth, quieter mistake. Panicking. Founders often abandon a working new site SEO plan at week six because clicks are still near zero, which is exactly when the data starts becoming useful. Give the plan its full quarter.

How do you measure progress when clicks are still zero?

Clicks are a lagging indicator on a new domain. Coverage state and crawl behaviour lead them by weeks. If you only watch clicks, you will spend two months staring at a flat line and learn nothing from it.

Track four things instead, in this order. Coverage state per URL, using the URL Inspection tool in Google Search Console, which tells you whether a page is on Google, not on Google, or entirely unknown. Crawl frequency, from your server logs or the Search Console crawl stats report. Impressions, which appear well before clicks. Then average position, read only as a trend and never as a single number.

Here is the practical rule I would apply. A URL that moves from unknown to crawled to indexed inside 14 days means discovery is working. A URL still unknown after 21 days means something is blocking it, and no amount of extra content will help. This is the part of new site SEO that separates people who fix problems from people who wait and hope. Impressions with a zero click-through rate are also fine at this stage, because ranking at position 45 earns impressions and almost never earns clicks.

Not indexed yet, or actually penalised?

New site owners confuse these constantly, and the fix for each is the opposite of the other. Slow indexing needs patience and better links. A penalty needs a specific repair and a reconsideration request. Guess wrong and you waste your entire quarter, because new site SEO work aimed at the wrong cause fixes nothing.

There is a third category people forget, and it only applies once you have some traction. A loss that lines up with a broad algorithm rollout is neither a penalty nor slow discovery. Our breakdown of why core updates are recalibrations rather than penalties explains what realistic recovery looks like. The single fastest check, though, takes ten seconds. Search your exact brand name in Google. If your home page shows up, you are not penalised. You are just early, and early is a solvable condition.

Does new site SEO change across the USA, Europe, and South Korea?

The mechanics stay the same everywhere. The competition and the search surface do not. A new domain in the United States faces the deepest field. European markets often have thinner English and local-language results, which makes early wins realistic. South Korea plays by partly different rules.

In Europe, a country domain such as a .de or .co.uk address still helps users trust a local business, and it signals regional relevance at almost no cost. Expect GDPR consent banners to dent your Google Analytics 4 numbers, so treat Search Console as the cleaner source. In Germany or the Netherlands, a well-written page in the local language often beats a translated one from a much bigger competitor.

South Korea is genuinely different. Google has grown into a serious force there, but Naver still leads on most Korean measurements of query volume, with Kakao's Daum holding a small share behind it, and the great majority of people search in Korean. If Korea matters to your business, plan a real Korean-language version rather than leaning on an English page, and register with Naver's own webmaster tools alongside Google. Pricing shown in KRW also converts better than a converted USD figure.

Frequently asked questions

How long does it take a brand-new website to rank on Google?

Plan for three to six months before meaningful organic traffic, and up to a year for competitive terms. Ahrefs found only 1.74% of new pages reach the top ten within a year, and about 41% of those that do arrive within the first month, after which the odds fall away sharply. Low-competition long-tail queries move much faster, sometimes within four to eight weeks. Your realistic day-90 outcome is full indexing, a stable crawl pattern, and a small number of long-tail rankings. Treat any traffic before that as a bonus, not the plan.

Is the Google sandbox real?

Not as a deliberate filter. Google has never confirmed one, and John Mueller has repeatedly said no traditional sandbox exists. What does exist is a new domain with no links, no click data, and a minimal crawl allocation, which produces the same slow start people describe. The distinction matters because it changes your response. You cannot wait out a filter, but you can earn links, mentions, and crawl frequency. Treat the delay as missing evidence rather than punishment and your new site SEO plan becomes obvious.

Should I buy an expired domain to skip the wait?

Rarely worth it. An aged domain can carry useful link equity, but it can equally carry spam history, a manual action, or topical baggage with nothing to do with your business. Check the Wayback Machine for what the site was, review the backlink profile in Ahrefs or Semrush, and confirm it is not blacklisted. If any of those look wrong, walk away. A clean registration costs roughly 10 to 15 USD, or about 9 to 14 EUR, per year, which is a small price for a known-good starting point.

How many pages should a new site launch with?

Between eight and fifteen substantial pages, plus the standard home, about, and contact pages. That is enough for Google to understand your topic without diluting crawl budget across thin content. Every page should answer a real question completely. Publishing 200 pages in month one on a domain with no trust is the most common self-inflicted wound, because it teaches Google that most of your site is low value. The quality of your first ten pages predicts the next ninety days better than quantity does.

When should a new site start building links?

Around month three, and only once you have something genuinely link-worthy. Before that, focus on being mentioned rather than acquiring links. Launch on Product Hunt, answer questions properly on Reddit, publish original data, or ship a free tool. Buying links in week two on a domain with zero traffic creates an obvious footprint. Natural early links tend to come from your own network, local business directories, industry associations, and suppliers. Those are slower and smaller, but considerably safer than any paid package.

Why is my new site indexed but getting zero clicks?

Because indexed is not the same as ranking. A page sitting at position 40 collects impressions and almost never earns a click, since most searchers never scroll past the first few results. Check your average position in Search Console. If you sit between 20 and 60, the page works but needs authority and a better intent match. If you rank in the top ten with no clicks, the problem is your title and description instead, and rewriting those is a fast, cheap fix.

Do I need a sitemap and Search Console on day one?

Search Console yes, absolutely. It is free, it is your only direct view of coverage state, and the data only accumulates from the day you verify it. A sitemap is useful but less critical than people assume, since Google mainly finds pages through links. Submit one anyway, because it takes minutes and gives you a clear error report. Add Bing Webmaster Tools at the same time, as Bing data increasingly feeds the AI assistants that cite sources.

Can I rank without any backlinks in the first 90 days?

Yes, for genuinely low-competition queries. Long-tail questions with weak existing results are winnable on content quality and clear structure alone. What you cannot do without links is compete for commercial head terms, where every page on page one holds hundreds of referring domains. So build your first quarter around specific questions rather than broad categories. Links then arrive more naturally once you own pages worth citing, which is a far easier pitch than promoting an empty domain.

How do I know if my new site is penalised?

Check the manual actions and security issues reports in Google Search Console first. If both are clean, you almost certainly are not penalised. Then search your exact brand name. If your home page appears, Google trusts you enough to rank you for something. A genuine penalty usually follows a visible pattern: real traffic that disappears sharply after a specific date. A site that never had traffic has not been penalised. It simply has not been discovered and evaluated yet.

Does a .com rank better than a country domain like .de or .co.uk?

Not inherently. Google treats a .com as global and treats country domains as a strong signal for that specific country. For a business serving Germany, a .de address helps local relevance and local trust. For a company selling worldwide, a .com stays more flexible. The real ranking factors sit elsewhere: content quality, links, and intent match. Pick the extension your customers expect to see, and never accept a strange extension purely because a shorter name was free.

Your first 90 days, in priority order

Remember that 1.74% figure from the opening. It is not a wall. It is a filter that removes everyone who quits at week six, publishes 300 thin pages, or spends the budget on links before the site is worth linking to. Avoid those three and your odds improve sharply, because most of the field disqualifies itself.

If you do nothing else, reduce new site SEO to these four steps in order. Verify Google Search Console on day one. Publish ten pages that fully answer a specific question. Watch coverage state and crawl frequency instead of clicks. Then, at month three, earn your first real mentions. Everything else is refinement.

My prediction for the next two years is that new domains will get discovered faster by AI answer engines than by classic search, simply because those systems reward clear answers over accumulated authority. Sound new site SEO in 2026 means writing for both at once. NasSeoTools will keep publishing on this as the data changes.

What is your day-90 target for the domain you just launched, and how will you know you hit it?