eCommerce
How to start an online store.
The build is the easy part now. What decides whether a store works is chosen before any of it: what you sell, what it costs to deliver, and what it costs to find a buyer.
- 10 min read
- Updated 21 August 2026
- No sign-up to read
The short version
- Check the unit economics before building anything at all.
- A 30% gross margin cannot absorb paid acquisition. Aim well above it.
- Shipping and returns are the costs that quietly sink first stores.
- Launch with a narrow range. Breadth is a problem you earn.
Start with the numbers, not the storefront
Most failed stores were doomed at the spreadsheet stage. Before choosing a platform, work out what one sale actually earns you:
- Product cost, landed — including shipping to you and any duty.
- Payment processing — roughly 2–3% plus a fixed fee.
- Shipping to the customer, whether or not you charge for it.
- Packaging.
- Returns — assume 5–10% for most categories, far higher for apparel.
What remains is your real gross margin. Run it through the profit margin calculator. Below about 40% you have very little room to pay for customers, and paid acquisition will not work.
Choosing what to sell
Whatever the category, the same characteristics keep recurring in stores that survive:
- Enough margin to pay for acquisition. This is the one that eliminates most ideas.
- Ships without drama. Light, sturdy, not perishable, not restricted.
- A reason to come back. Consumables and refills beat one-off purchases because the second sale costs nothing to acquire.
- Something specific to say. Reselling a product a hundred other stores carry leaves you competing purely on price.
Dropshipping, honestly
Dropshipping removes inventory risk, which is genuinely valuable when you are testing. It also gives you thin margins, no control over shipping times, and a product identical to everyone else's.
It works as a way to validate demand before committing capital. As a long-term model it is difficult precisely because the thing that makes it easy to start — no barrier to entry — applies equally to everyone who comes after you. If it works, plan the transition to holding your own stock or having something made.
Platform and payments
For most first stores a hosted commerce platform is the right answer: PCI compliance, payment integrations, tax handling and security are solved problems you should not be solving yourself. Custom builds make sense when you have genuinely unusual requirements, not before.
On checkout, the rules are well established and consistently ignored:
- Offer guest checkout. Forcing account creation is among the largest single causes of abandonment.
- Show the full price — including shipping — as early as possible. Unexpected costs at the final step are the most cited reason for abandoning a cart.
- Offer the payment methods your market expects, including whatever local option is normal where you sell.
- Ask for the minimum information the order requires.
Getting the first orders
A new store has no traffic and no trust. Both have to be built deliberately.
- Tell people you know. Unglamorous, and the source of most first orders.
- Pick one channel. One social platform, or one paid channel, or one marketplace. Learn it properly before adding another.
- Collect email addresses from day one. The one audience you own outright.
- Get reviews early. Ask every early customer. Nothing else converts strangers as reliably.
- Tag every link. Use the UTM builder so you can tell which effort actually produced sales.
When you start paying for traffic, judge it against your margin rather than revenue — the ROAS calculator shows the return you need to break even, which is frequently higher than people assume.
Do it now, free
The tools for this job.
Questions
People also ask.
Want the store built properly?
We build commerce that is fast, converts, and does not fall over on launch day.