Most of what determines whether an online store works is decided before a single page is built. These are the questions worth settling first, while changing your mind is still free.
An online store that disappoints usually disappoints for reasons that had nothing to do with the design. The decisions that decide whether it works are commercial, and most of them are cheaper to settle now, on paper, than to discover halfway through a build. Here are the ones worth answering first.
What are you actually selling?
A fixed catalogue of physical products, a handful of services quoted per job, digital downloads, subscriptions, or a mix — each behaves differently in a checkout. A subscription needs a way to renew and cancel. A quoted service cannot show a single "price" honestly. A digital download needs a delivery link, not a courier. Deciding the shape of what you sell tells you what the store has to do before anyone designs a page.
How does money reach you?
In Kenya this usually means M-Pesa, card, or both, and the answer changes what the checkout looks like and what it costs to run. It also raises a question people skip: what happens the moment after payment. Who confirms the order, who fulfils it, and how does the customer know it worked? A store that takes money but leaves the customer staring at a blank page has not finished the job it started.
The unglamorous half of ecommerce
The checkout is the part everyone thinks about. The part that actually determines whether customers come back is what happens after it — confirmation, fulfilment, and a clear answer when something goes wrong.
Who manages the stock and the orders?
Someone has to add products, update prices, mark things out of stock, and process orders as they arrive. If that someone is you, the store needs an admin area you can actually use without a manual. If nobody has time to keep the catalogue current, an out-of-date store is worse than none — customers order things you no longer sell, and the first they hear of it is an apology.
What is your delivery and returns position?
Decide, in writing, before launch: how orders are delivered or fulfilled, how long it takes, what a customer can return, within how long, and who pays return costs. These are not formalities. They are the difference between a dispute you can resolve by pointing at a published policy and one that turns into an argument you cannot win. A store without clear policies is a store that has decided to settle every disagreement from scratch.
Can a stranger trust it?
A first-time visitor is deciding whether to hand you money on the strength of a few seconds. The things that earn that trust are unglamorous and specific: a real business name and contact details, prices that are clear rather than "contact us", policies that exist and read like a real business wrote them, and a checkout that looks and behaves like the ones they already use. None of it is expensive. Its absence is what makes a good product sit in an abandoned cart.
Start smaller than you think
It is tempting to launch with the full catalogue, every payment method and every feature. A store that launches with a tight range you can fulfil well, and grows as demand shows you what to add, beats one that launches broad and cannot keep any of it current. You can always add. Taking things away after customers have seen them is harder.