how to start an online store from scratch
the store is no longer the hard part. software builds you a working shop in an evening - which is exactly why thousands of empty ones open every day, polished and unvisited.
so the order of operations is the whole game. decide what you sell and who buys it before touching a platform, and the storefront becomes an evening of assembly instead of a monument to a decision you skipped.
the offer comes first
three questions, on paper, before any signup:
- what, exactly? one product or a tight family - not a category. "linen aprons for home bakers" opens doors; "kitchen goods" opens none. specific stores get found, remembered, and recommended; general stores compete with everyone about nothing.
- who wants it already? name the buyer the way a writer names a reader: one person, one need, searching real words. if you cannot say what they type into a search box, the store has no front door yet.
- why yours? sourced better, made by you, curated with taste, explained properly - one clear reason. without it the only lever left is price, and the race down is already lost.
a store that can answer all three in three sentences is ready to build. worked example for the rest of this post: maya again, the shop owner from the automation post - selling hand-poured candles to people who buy gifts under $40.
pick the ground
the tradeoffs straight: a hosted platform is the fastest reliable start, and its costs scale with you forever - rent plus a slice of every sale. your own stack costs more attention up front and then belongs to you: the hosting, the customer data, the exit. it is the store version of owning the asset.
the lean by case: testing whether anyone buys at all - hosted, cheapest plan, decide later. building the long-term thing on a budget - your own stack, one evening more. maya wants the second: her margins are thin enough that percentage fees sting.
build night, step by step
product pages that sell without shouting
the copy rules are the site's rules: say the thing. what it is, in the buyer's words. who it is for, named plainly ("burns about 40 hours; the right gift when you need under-$40 and thoughtful"). what to know before buying - size, materials, care, the caveat you would tell a friend. photograph in daylight against one background.
no urgency theater, no fake countdown timers, no "only 2 left" that lies. trust converts better than pressure, and it compounds instead of burning out.
the first ten visitors
marketing is its own craft, but the store's first week has a checklist: tell the people who already know you, once, plainly. put the link where your buyer already looks - the relevant community, answered questions, a marketplace listing that points home. start the store's email list on day one, because the audience you can reach directly is the asset the store builds after the products. paid ads wait until a stranger has bought - ads amplify what works, and nothing is proven yet.
your hour
- answer the three offer questions in writing: what exactly, who already wants it, why yours
- pick the ground with one sentence of reasoning: testing (hosted) or building (own stack)
- buy the domain tonight - the decision that makes it real
- book build night in the calendar and prep the five product photos before it arrives
faq
how do i start an online shopping site from scratch?
offer first: one specific product family, a named buyer, one reason yours. then a build evening: domain, platform or wordpress + woocommerce, five products, payments, trust pages, one test order.
what does it cost to start an online store?
your own stack: roughly a domain plus a few dollars a month of hosting. hosted platforms: a monthly plan plus transaction fees. inventory is the real variable - which is why five products beat fifty to start.
shopify or my own website?
hosted to test fast, own stack to own the result - lower running costs, your customer data, your exit. the decision mirrors every rent-or-own choice on this site.
more in the notes.