A new website plays by different rules than one that has been running for three years, and half the standard SEO advice simply does not apply to it. A new domain has no accumulated trust, no behavioural history, and not a single page it is already known for. So the order of work has to change: first what makes the site visible to crawlers, then what earns the first entries, and only then the things people usually want to discuss first. Below is a plan for the first three months, tested on projects we launched from zero.

What has to be done before launch

The most expensive mistakes on a new site are made before it becomes reachable. They can be fixed later, but it costs months: re-indexation, consolidating duplicates and moving URLs are always slower than getting it right the first time.

Month one: make the site visible

The goal of month one is indexation, not rankings. Until pages are in the index, work on copy has nowhere to show up. It is the least impressive stage and the one whose omission stalls everything else.

1. 1Verify the domain in Search Console and submit the sitemap. Then check the indexing report weekly rather than daily — its data updates with a delay.
2. 2Submit key pages for indexing manually. This does not speed up ranking, but it cuts time to first crawl from weeks to days.
3. 3Check how a crawler sees your pages. If content only appears after scripts execute, part of it may never reach the index.
4. 4Remove duplicates immediately. Pagination, filters, sort parameters and index.html variants create copies that dilute signals before any traffic exists.
5. 5Lock in baseline speed. Fixing Core Web Vitals across eight pages is far cheaper than across eighty.

Month two: first pages built around demand

The typical mistake is starting with twenty blog articles. A new site needs commercial pages first: they answer queries with buying intent, there are fewer of them, and they produce the first enquiries. The blog works, but its effect is shifted by roughly half a year, and that is normal.

Keyword research at this stage is not exhaustive but minimally sufficient: two or three clusters where you are genuinely strong. The step-by-step method is in our keyword research guide, and which pages should exist at all is covered in the piece on website structure.

Month three: external signals and first conclusions

By month three the first data arrives: impressions, a handful of queries, occasionally positions in the twenties or thirties. This is the moment to adjust the plan against facts rather than assumptions. In parallel, start the simplest external work — industry directories, partner mentions, company profiles. This is not link buying; it is the baseline presence without which a new domain looks isolated.

What not to do in the first months

• Do not change URLs after indexation. Every address change resets what accumulated and requires redirects.
• Do not publish dozens of thin pages. On a new domain that is a direct route to "crawled — currently not indexed".
• Do not buy links in bulk. A young domain with no history is the worst candidate for an aggressive profile.
• Do not judge results by rankings in month two. Realistic timelines are covered in our piece on how long SEO takes.
• Do not start a redesign mid-process. It writes off part of the work.

Frequently asked questions about SEO for new sites

In short

• Month one is about indexation, not rankings — until pages are indexed, other work has nowhere to show.
• The two classic launch failures, a leftover noindex and a blocking robots.txt, hit nearly every second project.
• Four to eight strong pages are enough at launch; dozens of thin ones hurt a new domain.
• Commercial pages in months one and two, blog from month three with a 4–8 month lag understood upfront.
• Success on day 90 is indexation, rising impressions and green Core Web Vitals — not traffic.
• Fix URLs, structure and design before indexation: changing them afterwards costs months.