Starting an Online Store: What to Decide Before You Build

An online store is not a website with a buy button. It is a small logistics operation that happens to have a website attached, and most of what determines whether it works is decided before anyone writes a line of code.

People discover this in the wrong order. The site gets built, and then somebody asks how returns work, and there is a silence.

Here is what to settle first.

1. How does the money actually reach you?

Not "we'll take card payments". Specifically: which provider, in which country, into which account.

Payment processors have requirements. They want a registered business, sometimes in a specific country. They have different fees, different payout schedules, different rules about what you are allowed to sell. Some will not work with certain industries at all, and you find this out during application, not before.

Settle before building:

• Which provider, and have you confirmed they accept your business type and country?
• What are the fees, per transaction and monthly? On thin margins this decides whether the store works at all.
• How long until money lands in your account? Two days and thirty days are different businesses.
• What happens on a chargeback? Who eats it? If your business is not registered yet, sort that first. Every serious payment provider wants a real entity, and no amount of good website makes that go away.

2. What happens after someone clicks buy?

The order arrives. Now what — physically, in the real world, by a person?

Somebody prints a label. Somebody puts the thing in a box. Somebody takes it to a courier. Somebody updates the order so the customer gets a tracking number. If that somebody is you, you have just added a job to your week, and at ten orders a day it is most of a job.

Settle before building:

• Who packs and ships? You, a warehouse, a fulfilment service?
• Which courier, and does the site need to talk to them or will you copy numbers by hand?
• How fast do you promise? Whatever you write on the site is a promise, and people screenshot it.
• What is the cost, and who pays it? Free shipping is not free; it is priced in and you should know where. The website can automate a lot of this. It cannot pack a box. Decide who does before deciding what the site does.

3. Returns

Nobody wants to think about returns before the first sale. Then the first return arrives and there is no process, and you invent one badly while a customer waits.

In the EU, distance selling comes with legal return rights — a customer can generally send it back within a set window without giving a reason. This is not optional and it is not negotiable by writing something different on your website.

Settle before building:

• What is your return window, and does it meet the law where your customers are?
• Who pays return postage?
• Refund to card, or store credit? What is the legal position?
• What condition must it be in?
• Who inspects it, and how does the money go back? This becomes text on your site, an email template, and a process. All three need to exist on day one.

4. Tax, which is worse than you think

If you sell across borders, VAT gets complicated fast, and getting it wrong is expensive in a way that compounds quietly.

Selling from one country to consumers in another means rules about registration thresholds, where the tax is owed, what your invoices must show. The EU has a scheme to simplify this. It is still not simple.

Settle before building:

• Which countries will you sell to? Restricting to two is a legitimate strategy and much cheaper.
• Are prices shown with or without tax? This differs by market and is decided in the design, not afterwards.
• Do you need to register anywhere?
• What must an invoice contain? Talk to an accountant before the site is built. Not after. The answers change what gets built — how prices are stored, what the checkout collects, what the invoice looks like. Retrofitting this is genuinely painful.

5. Where does stock live?

You have twelve of something. Someone buys one. Now you have eleven — everywhere that number appears.

If the site is the only place you sell, this is easy: the site knows. If you also sell in a shop, on a marketplace, at markets, then "how many do we have" has several possible answers and they disagree with each other. Somebody sells the last one twice, and now you are emailing a customer to say sorry.

Settle before building:

• Is the website the single source of truth for stock, or is something else?
• If something else, does the site need to talk to it, or will someone update by hand?
• What happens when stock hits zero — hide it, show it as sold out, allow backorder?
• Who updates it, and how often, honestly?

6. How many products, and how do people find them?

Twelve products and twelve hundred are different websites.

At twelve, a list works. Everyone sees everything. Search is unnecessary.

At twelve hundred, the entire product is findability — categories, filters, search that tolerates typos, sorting. Get this wrong and people leave, because a customer who cannot find the thing does not buy the thing. This is most of the design work, and it is invisible in a mockup.

Settle before building: roughly how many now, roughly how many in two years, and how does a customer who does not know your product names find what they want?

7. What makes anyone trust you with a card number?

This is the part that gets skipped, and it decides more sales than the design does.

A stranger is about to type card details into a site they found nine minutes ago. Every unanswered question is a reason to close the tab.

• Is there a real address and a phone number that works?
• Do returns and shipping have their own clear pages, not buried in a footer link?
• Are there photographs of the actual product, from angles a real person would want?
• Is there any evidence other humans have bought from you?
• Does the total show before the final step, including shipping? That last one matters enormously. Surprise costs at checkout is the single most reliable way to lose a sale that was already made. The customer has decided, they are entering details, and then the number changes. They leave, and it feels like a trick even when it was not.

Trust signals are what separates a site that gets visits from a site that gets orders — and on an e-commerce site the stakes are higher, because you are asking for money rather than an email.

The order to do it in

1. Register the business. Payments require it.
2. Choose a payment provider and confirm they accept you. Before anything else.
3. Talk to an accountant about tax. Before the site is designed.
4. Decide who packs and ships, and how fast you promise.
5. Write the returns policy.
6. Decide the stock question.
7. Then talk to someone about the website. Steps 1 through 6 have nothing to do with web development and they determine what gets built. A studio that starts asking about payment providers and tax before showing you designs is not stalling — that is what competence looks like on this kind of project.

The honest note about cost

An online store is the most expensive thing a small business commonly commissions, and the price range is wide. The gap between the cheap end and the expensive end is real work, most of it invisible: what happens when a payment half-fails, when stock runs out mid-checkout, when someone orders from a country you did not plan for.

The cheap version handles the happy path. Every edge case becomes a person emailing you, and each of those emails is your time, at a price.

We're DAKER, a web design and development studio in Baku working with startups and small businesses across Europe. We build online stores from scratch, with an admin panel that fits how you actually work — but we'll ask you all of the above before we quote, because the answers change the project. Get in touch.