The first week after your shop goes live

A shop that has been tested is not the same as a shop that has been used. Real customers do things no test covers, on devices and connections you do not have, and the first week is when that shows. It is also the week most owners spend redesigning things instead of watching what is actually happening.

Why follow the first orders all the way through?

For the first handful of orders, watch each one end to end rather than glancing at the total. Did the confirmation email arrive, and did it arrive quickly? Did the address come through complete? Did the delivery option the customer chose match what you can actually do?

This is tedious and it is the highest-value thing you will do that week. Faults in this chain are silent by nature: the customer sees a problem and leaves, and nothing in your admin records that anything went wrong.

Keep the first orders somewhere you can look at them again. When something odd happens in month two, the early ones are the reference for what normal looked like.

Which faults do only real customers find?

Real people put spaces in card names, use addresses that do not fit the form, order from a phone with a flaky signal, and pick the delivery option you did not expect anyone to choose. Each of those finds an edge no test does.

Watch particularly for anything that fails at the last step. A payment that declines for an ordinary reason should return the customer to a page that explains what to do next, not to a dead end. Test one declined payment deliberately if you have not.

Also check what happens when something is out of stock. Selling an item you cannot ship is a bad first impression, and stock settings are one of the most common things to be left wrong at launch.

Why write down every question customers ask?

Early customers ask more questions than later ones, because there is less about you to reassure them. Answering quickly matters more than answering perfectly.

Keep a list of every question. That list is the most valuable thing launch week produces: it tells you exactly which sentence is missing from a product page, and each question answered on the page is one you never have to answer again.

If the same question comes up three times, it is not a question. It is a fault in the page, and it is costing you the buyers who did not bother to ask.

How do you confirm the money actually arrives?

An order in the admin is not money in the bank. Check that the first payouts actually reach your account, that the amounts make sense after fees, and that refunds work when you issue one.

Issue a small refund on purpose in the first week. It is far better to discover a problem with refunds on an order you control than on a customer who is already unhappy.

Check the tax settings against a real order too. Getting this wrong quietly for months is much harder to correct than getting it right in week one.

Should you redesign anything in launch week?

The urge in launch week is to change the look. Almost always this is nerves rather than evidence, and it destroys the only clean read you will ever get of how the shop performs as built.

Fix faults immediately: broken things, wrong prices, missing information, anything that misleads. Leave preferences alone. The difference is whether a customer would be harmed by it or you simply do not like it.

Give it enough orders to mean something before you change anything on purpose. Early numbers are noisy, and a shop with a handful of sales cannot tell you whether a change helped.

What should you measure in week one, and what should you ignore?

Worth watching in week one: orders that complete, questions asked, anything that errors, and whether the shop works on a phone on mobile data. All four are about correctness, which is the only thing you can genuinely judge this early.

Not worth watching yet: conversion rate, bounce rate, and any ranking. There is not enough data for those to mean anything, and staring at them produces confident conclusions built on almost nothing.

Search engines will also take their own time to find and index a new shop. That is normal, it is not a fault, and nothing you do in week one meaningfully speeds it up.

What to do next

Spend launch week watching rather than changing. Follow the first orders end to end, issue one refund on purpose, keep a list of every question you are asked, and fix only what is genuinely broken. Then leave the design alone until you have enough orders to learn from. If you would rather somebody else watched it, that is what a Run plan is for — Run Core is £29/month and there is no minimum term.

Common questions