What to do first, in order, on a site with nothing
You have a new site, a long list of advice, and no idea which part to start with.
Short answer
Confirm you can be indexed, publish one page that answers one specific query better than what currently ranks, make sure that content is in the HTML, link to it from your homepage, then wait and read impressions. Everything else, including most technical SEO and all link building, comes after you know the machine works at all.
Move one: prove you can be indexed
Search your exact domain. If nothing comes back, nothing else matters and you have a crawlability or noindex problem to fix today. If it does come back, you are indexed and your problem is ranking, which is an entirely different set of work.
Move two: one page, one query
Not ten pages. One, aimed at a query you have seen a real person ask, answering it more completely than anything currently on page one. If a single page cannot collect impressions, ten will not either, and you will have spent the month finding that out expensively.
Move three: put the content in the HTML
If the page is assembled in the browser, the first crawl sees a shell, indexing is deferred, and every crawler that does not run JavaScript sees nothing. Server-render or prerender the pages that need to rank. This is a one-time cost that removes a whole class of problem.
Move four: link to it
A page nothing points at is a suggestion. Link it from your homepage and from any related page. Discovery, crawl frequency and importance all follow internal links, and this is the cheapest lever a new site has.
Move five: wait, and watch impressions
Weeks, not days. Impressions are the first thing to move and the only honest early signal. If impressions climb, the machine works and you are early. If they stay at zero for a month, go back to move one, because something is broken rather than slow.
What to ignore for now
Link building, before you have anything worth linking to. Schema, before you rank. Core Web Vitals, before you have traffic to lose. Content calendars, before you know which single page worked. Each is legitimate later and each is a way to feel productive now.
Then repeat, in small batches
Add pages a handful at a time and check impressions per new page before adding more. When that number falls, stop adding and improve what exists. This one habit prevents the long tail of dead pages that most sites accumulate.
Questions
- How many pages should I have before expecting traffic?
- One that genuinely answers a query beats fifty that approximately do. The number is not the variable; whether any single page is the best available answer is.
- Should I set up analytics first?
- Set up Search Console before you publish, because it only collects data going forward and you cannot backfill. That is the one setup task worth doing before anything else.
- When do I start on links?
- Once you have something worth linking to. Outreach for a page that is not the best answer to anything is the least productive work in the discipline.
Measured, not asserted
This is the order ihatepdf.cv actually ran, reaching 100,542 users with $0 spent on advertising. It also shows the cost of not batching: of 160 pages, 93 of 160 earned fewer than ten clicks, and 75.4% of all traffic came from five.
Free tool for this: Crawler View. No account, nothing uploaded.
Where this goes deeper
Every number on this page comes from one complete dataset: one product taken from zero to 100K+ users on search alone, with nothing spent on advertising. The full argument is Chapter 3 and Chapter 7 of the book. Five chapters are free to read.