The First 90 Days of SEO for a Brand New Site
Most of what people do in month one is either premature or pointless. Here is the short list that isn't.
· 8 min read
New sites are slow. Not broken — slow. Google has no history with your domain, no reason to trust it, and no evidence that crawling it is a good use of its time. It will take weeks to form an opinion, and almost everything people do in that window to hurry it along is wasted motion.
Here is what is actually worth doing, in order, and what to ignore.
Week 1: the foundations you cannot retrofit cheaply
Verify a domain property, not a URL-prefix property
A domain property covers every subdomain and protocol at once. Do this first — a URL-prefix property is how people end up with all their real traffic sitting in a property they forgot they created. See verifying site ownership.
Pick one hostname and commit to it everywhere
Apex or www. Pick one. It must match across your sitemap, your canonical tags, your internal links, and your host's redirect direction — and the redirect must be a 301 or 308, never a 302 or 307. A temporary redirect tells Google to keep the *other* URL indexed, which is the opposite of what you want.
Check the canonical on a page that is not the homepage
curl -s yoursite.com/about | grep canonical. The href must be that page's own URL. If it is your homepage, you have the single most expensive bug in this list, and it takes one line to fix.Submit a sitemap
Generate it from your content, not by hand. Submit it once. You never need to submit it again — the ping endpoint was deprecated in 2023 and resubmitting does nothing.
Weeks 2–4: the silence, and why it is normal
You will check Search Console every day and see nothing. No impressions, or three impressions. Pages sitting in Discovered – currently not indexed. It feels exactly like a bug.
It is not a bug. It is Google deciding, slowly, whether your domain is worth crawling. New domains have no authority, so they get a small crawl budget, and the queue moves at a pace that has nothing to do with how good your content is.
What genuinely helps, in rough order of value:
- A fast server. Google throttles crawling on slow sites to avoid degrading them. TTFB is a crawl-rate input, and it is one of the few things fully in your control.
- Internal links. An orphaned page is a page Google has little reason to prioritize. Every page should be reachable from your nav, your footer, or a hub page.
- Any real external link. One link from a site Google already trusts does more for crawl priority in month one than everything else on this list combined.
- Fewer, better pages. Twenty thin pages compete with each other for a crawl budget that will not stretch. Five good ones will be crawled sooner and judged better.
What does not help: requesting indexing on the same URL repeatedly, resubmitting the sitemap, tweaking changefreq and priority (Google ignores both), and refreshing the report.
Weeks 4–8: the first real signal
Impressions start appearing. They will be for queries you did not target and did not expect, at positions in the twenties and thirties. This is the most useful data you will get all quarter, and it is routinely dismissed because the numbers look trivially small.
Do not dismiss it. Those queries are Google telling you what it currently thinks your site is *about*. That opinion is what everything else gets built on.
- Queries you did not expect, ranking surprisingly well → Google thinks you are credible here. This is the cheapest opportunity you will ever get. Write more about it.
- Queries you targeted, nowhere to be found → your page is not competitive, or Google has not indexed it. Check which, because the two have completely different fixes.
- Impressions but no clicks at all → normal at position 25. Nobody clicks position 25. Ignore CTR until you are inside the top 15.
Weeks 8–12: the first thing worth optimizing
By now something will have drifted into the 5–15 position range with real impressions behind it. That is your first genuine optimization target, and it is the one that pays fastest.
You already rank. Google already considers you relevant. People are seeing you and choosing someone else. Rewriting the title and meta description to match what those people were actually searching for is an afternoon of work, requires no new content and no links, and is the highest-return move available to a young site.
Full method: Clicks, Impressions, CTR & Position Explained.
What to ignore for the entire 90 days
| Thing | Why not yet |
|---|---|
| Core Web Vitals | There is not enough traffic to generate field data. And it is a tiebreaker, not a lever. |
| Buying an SEO tool | You have no data to analyze and no competitors to study yet. Search Console has everything you need in month one. |
| Keyword density, LSI, and similar folklore | Not a thing. It has not been a thing for over a decade. |
| Daily rank checking | The data lags 1–3 days and moves on noise. You will learn nothing and lose an hour a week. |
| Link building outreach | Before you have anything worth linking to, this is cold-emailing strangers to ask a favour with nothing to offer. |
The one habit that matters
Check the Page indexing report weekly. Not the Performance report — the indexing one.
Everything in SEO is downstream of being indexed, and indexing problems are silent, gradual, and cumulative. A canonical bug caught in week one costs you a deploy. The same bug caught in week twelve has already eaten your quarter, and you will spend the next one waiting for Google to re-crawl its way back to trusting you.