The short version
- A new domain is not penalised. It is simply unknown - and that distinction changes the whole plan.
- The first ninety days are for being discovered and gathering data, not for ranking on head terms.
- Eight solid pages beat thirty thin ones: on a new domain, dilution costs double.
- The first real signal is impressions, not clicks. They arrive between week three and week six.
- Without paid search in the early months you have no data on what converts, and you will pick queries on instinct.
Launching a business and putting a new website online puts you in an awkward position: everything written about SEO assumes a site with history. Analyse your impressions, look at declining pages, compare with last year. With a domain published three weeks ago, none of that advice applies.
This article is about exactly that phase. What actually happens in the first ninety days, what is reasonable to expect month by month, which mistakes cost a full quarter, and in what order to do things when you have no data at all.
A new domain is not penalised
Worth clearing up a misconception that has circulated for years. There is no penalty for new domains. What exists is more mundane and far more useful to understand: a freshly published site has not yet been discovered, evaluated and placed, and those three steps take time.
The practical difference is substantial. If you believe you are penalised, you wait for it to pass. If you understand you are unknown, you work on discovery — which is something you can act on directly.
| Missing element | How it is built | Typical timeframe |
|---|---|---|
| Discovery | Sitemap, URL submission, links from already-crawled sources | Days |
| Indexing | Content that deserves a place in the index | 1-4 weeks |
| Crawl frequency | Regular, useful publishing | 1-3 months |
| Authority | Citations and links from real sources | 6-18 months |
| Brand recognition | Activity away from the website | Ongoing |
In the first three months you can realistically achieve discovery, indexing and early positions on the very specific long tail. You do not get the sector's head terms, and no serious supplier promises them.
Month 1: getting found
The first month produces no traffic, and should not. It produces the conditions that make traffic possible.
The technical base, verified rather than assumed. The site returns 200 on the pages that matter, there is no noindex inherited from staging — the single most frequent launch error — every page's canonical points at itself, and the sitemap contains only indexable URLs and has been submitted.
The first eight to ten pages, not thirty. One page per service people search for separately, a contact page, and two or three pages answering the questions customers ask before buying. On a new domain, publishing thirty thin pages is worse than publishing eight solid ones: it dilutes the attention the engine gives you and fills the index with material that competes for nothing.
Local presence. Verified business listing, correct primary category, hours, real photographs. For a city business this channel produces results in weeks while organic takes months: it is the fastest source of customers available at the start.
Measurement, from day one. This is the step almost everyone defers and later regrets: without a baseline you will never know what worked. With Search Console analytics connected on day one you accumulate the history that in six months will be the only data you have.
Month 2: confirming the engine can see you
The second month is diagnostic. The question is not yet "how much traffic" but "are the crawlers coming?".
Here the bot visit log answers directly a question that is otherwise approached by guesswork. The three possible situations and what they mean:
No visits
A discovery problem. No sitemap, no internal links, or no link at all from an already-known external source.
Visits but no indexing
The content was seen and declined. Not a technical problem: a value problem.
Indexed, zero impressions
You are in the index but compete for nothing. The intent was wrong, not the writing.
Three diagnoses, three different actions. Without this data the typical reaction is to rewrite the content, which addresses only one of the three.
In parallel, month two is when newly published pages should be submitted for indexing and then checked to confirm they were actually crawled. It does not force a ranking, but it removes the waiting days in the discovery phase — which on a domain with no history is the slowest part of the process.
Month 3: the first real signals
Between week six and week twelve impressions start arriving, and this is where the project stops being theoretical.
The most valuable data at this stage is not clicks: it is the queries appearing that you never chose. A new site always starts surfacing for phrasings nobody planned, and those are the most reliable guide to what to write next — because they are real demand in your market rather than a tool's estimate.
The correct reading at ninety days: rising impressions, however small, means the project is working. Zero impressions with indexed pages means you wrote for the wrong intents. Zero indexing means the problem sits upstream and no amount of content will solve it.
The first eight pages: which ones, exactly
"Eight solid pages" stays vague until you say which. For a local business starting from zero, this is the structure that works in the large majority of cases.
One page per service, not one "services" page. If you offer four things people search for separately, you need four pages. A single page listing them all competes for none of the four searches, and on a new domain you have no authority to waste on pages that half-do everything.
A contact page that is actually a page. Address, tappable phone number, hours, map, area served and how to get there. On a local business it is often the second most visited page and it gets treated as a form.
Two or three pre-purchase question pages. What it costs, how long it takes, what is included and what is not. These answer the objections and they are also the only pages that can rank quickly on a domain with no authority: competitors rarely write them, because they prefer not to discuss prices.
An "about" page with real people. Names, photographs, experience, and credentials where the sector requires them. On a new site nobody has heard of, it is the one element substituting for the reputation you do not yet have.
What you do not need in the first three months: a blog. Publishing articles while the commercial pages do not exist yet is the wrong order, and it spreads the scarce authority you have across content that does not convert. A blog makes sense from month four, once the pages that must receive the visitors are standing and can receive the internal links.
Why paid search matters in the early months
It sounds off-topic in an SEO article and it is the most useful recommendation for anyone starting from zero.
A new site has two problems at once: no traffic and no data. The second is worse than the first, because it forces you to choose which queries to invest months of work in without knowing which ones produce customers.
Two weeks of a paid campaign across twenty candidate queries solves exactly that. It costs little, answers in days rather than quarters, and tells you which of those twenty generate enquiries at an acceptable cost. Content then gets written only for the winners, and the rest never consume three months of editorial work.
There is a secondary benefit nobody mentions: the search terms report shows the actual phrasings people used, including ones no keyword tool would have suggested. In a local market that is frequently where the most profitable intent turns out to be.
Paid first, to learn what converts; content second, on the queries that converted. Reversed, you write thirty pages and discover a quarter later that twelve were pointless.
Five mistakes that cost a quarter
Publishing everything on day one. Forty pages uploaded together on a domain with no history get crawled slowly and evaluated poorly. Eight pages then two or three a week is better: it also creates the regularity signal that increases crawl frequency.
Buying links in the early months. A new domain with a sudden, unnatural link profile is the case where the risk is most concrete. In the first six months the only sensible sources are the real ones: associations, suppliers, chambers of commerce, sponsorships you already pay for.
Changing the structure after a month. A new domain has no resilience to absorb a URL reorganisation. Structure is decided before publishing, then left alone for at least a year.
Judging at thirty days. At one month there is nothing to judge. The first sensible assessment is at ninety days, and it concerns impressions and indexing, not revenue.
Not collecting the history. The quietest mistake. Anyone who connects measurement tools in month six has permanently lost the first five months of data — which is exactly the data needed to understand what worked.
What to expect across twelve months
- Months 1-3: discoveryMain pages indexed, first impressions, first clicks on the very specific long tail. The local pack, where applicable, is already working.
- Months 4-6: commercial tailLong, specific queries enter the top positions. The first enquiries attributable to organic arrive. This is when the project becomes credible internally.
- Months 7-9: mid-difficulty intentMid-competition queries start moving. Volume becomes sufficient for meaningful analysis and for deciding where to go deeper.
- Months 10-12: head termsThe main queries start appearing. Rarely on the first page in a competitive market, but the trajectory becomes readable.
Anyone promising head terms in three months on a new domain, in a market with real competition, is either selling smoke or counting keywords nobody searches for.
Connect measurement on day one
Search analytics, rank tracking and a bot log: free, with no time limit. On a new domain, the history you start collecting today is the only history you will have in six months.
Sign in to Semalt See Fast IndexingFrequently asked questions
Is there a penalty for new domains?
No. There is the fact that the site has not yet been discovered, evaluated and placed. Three steps that take time, not a sanction.
How many pages should I publish initially?
Eight to ten solid ones, then two or three a week. Regularity matters more than the initial volume.
When will I see the first results?
Impressions between week three and week six, first long-tail clicks in months two to three, attributable enquiries from month four to six.
Should I buy links to speed things up?
No, and on a new domain it is when the risk is highest. The real sources you already have - associations, suppliers, sponsorships - are worth more and cost nothing.
Do I really need paid search?
If you can afford it, yes - not for the traffic but for the data. Two weeks of testing tells you which queries are worth writing for and saves a quarter of wrong content.
Conclusion
The first ninety days of a new domain are not judged on revenue but on three indicators: pages are being crawled, they enter the index, they start collecting impressions. If those three signals are present, the project is on the right trajectory even though the phone is not ringing yet.
The most important thing to do today, if the site is newly launched, is also the cheapest: connect measurement and start accumulating history. In six months that data will be the only thing letting you understand what worked — and it cannot be recovered retrospectively.
If you would like the first-ninety-days plan built with you for your sector, get in touch: the initial audit is free and delivered within 24 hours, even on a site published last week.
