In short
Work in this order: make the store technically visible, build the category structure from search demand rather than from your supplier's catalogue, get the product data right once so every downstream channel inherits it, then earn the first mentions from people who genuinely have reason to link to you. A new domain has no history, so expect the first two quarters to show indexing and impressions rather than revenue, and judge progress on whether the right pages are being found and crawled.
Get the foundations right while changing them is still cheap
Some decisions are almost free before launch and expensive afterwards. Choose the domain you intend to keep, decide between subfolders and subdomains for any second market, and settle the URL patterns your platform lets you control. Changing these later means redirect mapping and a period of instability.
Make sure the site can be crawled the moment you want it to be. New stores routinely launch with the staging noindex still in place, or with a robots file inherited from a development environment. Check the live site's robots rules and the head of a product page on day one, and inspect a live URL to confirm Google sees it as indexable.
Set up measurement before traffic arrives rather than after. Analytics with ecommerce events, Search Console verified with the sitemap submitted, and Merchant Center connected if you intend to sell through Shopping. Without these, the first months produce no evidence and you will be guessing.
If you are moving from an existing site or marketplace presence, map redirects from every old address before switching. This is the single most common source of lost traffic in a relaunch, and it is entirely preventable.
Build the category structure from demand
Most new stores organise their catalogue the way their supplier organises it, which reflects manufacturing rather than shopping. Search data lets you organise it the way people ask for things instead, and that structure is the skeleton everything else hangs on.
Start with the terms that describe what you sell, then look at the modifiers attached to them: type, use case, material, size, compatibility, price band, occasion. The modifiers with real volume become subcategories. The rest become filters. Do this before writing a single product description, because the structure determines which pages exist to write for.
Be realistic about which terms are reachable. A new domain competing for the broadest term in a mature category will wait a long time. The narrower, more specific pages are where a new store can get traction, and they also attract shoppers who know what they want.
Give each page a clear primary term, a distinct introduction, and links from and to its neighbours. Then stop expanding. Ten well-built category pages beat sixty empty ones, and a small site that looks finished performs better than a large one that looks unfinished.
Do the product data once, properly, and let every channel inherit it
Product attributes are shared infrastructure. The same fields feed your product pages, your structured data, your Merchant Center feed, your marketplace listings and any assistant that reads your catalogue, so cleaning them once pays several times.
The fields that matter most are the identifiers and the physical facts: brand, GTIN where the product has one, MPN, dimensions, weight, materials, colour, size, compatibility and country of origin. Get them into structured fields rather than into prose, then generate the prose from them.
Write product titles that name the brand, the product and the attribute a shopper searches with, and keep them consistent across the catalogue. Add Product structured data with price and availability that match the visible page, and check a sample in the rich results tooling.
Images deserve the same discipline: a consistent main shot on a plain background for Shopping surfaces, additional angles and in-use shots for the page, correct file naming and honest alt text. On a new store this is the cheapest time to establish the standard, because there are only a few hundred products rather than a few thousand.
Earn the first mentions, then measure the right things
A new domain has no links, and the ones that come easily are the ones nobody values. The routes that work early are relationships you already have: suppliers and brands with stockist or where to buy pages, trade associations and certification bodies you belong to, local and regional press if there is a genuine story, community groups and events you actually support, and specialists in your category who will write about a product if you send them one.
What should be avoided is anything bought at scale. Paid links, private networks and mass directory submissions fall under Google's spam policies, and a new domain has the least tolerance for that kind of pattern.
For the first six months, judge the program on leading indicators. Are your category and product URLs indexed. Is the number of queries you appear for growing. Are impressions rising on the pages you built deliberately. Is average position improving on the narrower terms even while the broad ones sit still.
Revenue from organic search on a new store tends to appear as a trickle that becomes a stream somewhere after the second or third quarter, and it appears first on the specific pages rather than the competitive ones. If the leading indicators are flat after two quarters of consistent work, the problem is usually that the pages are too similar to what already ranks rather than that more time is needed.
Related questions
Should a new store buy an aged domain to start faster?
It carries risk that is hard to assess from outside. History attached to a domain includes any penalties, spam signals and unrelated topical associations it accumulated, and a domain that changes subject entirely does not carry its previous standing across. Most new stores are better served putting the same money into product data, photography and content.
Does a new online store need a blog from day one?
No. Category and product pages come first because they sit on the queries that convert. Editorial content becomes worthwhile once those pages are complete and you have identified pre-purchase questions your commercial pages cannot answer honestly. Starting with guides while the collection pages are empty spreads thin content across more URLs.
How many products does a new store need before SEO is worth doing?
Any number, because the foundational work is the same at ten products or ten thousand. What changes is emphasis: a small catalogue depends heavily on the depth of each page and on being genuinely distinctive, while a large one depends on structure, templates and clean attribute data. Neither needs to wait.