How to Run an Online Store Day to Day: An Operations Guide

What are the steps to run an online store day to day?

Store operations is everything that happens between a visitor arriving and a happy customer coming back: the product data, the checkout, the payment, the parcel, the return, the reply and the numbers. Most of it is routine, which is good news, because routines can be written down, shared and checked. Here is the loop, in the order the work flows.

• 1. Assign owners: catalogue, orders and dispatch, returns, customer service, store health and the weekly numbers each need one named person.
• 2. Work out contribution per order, so you know what each order leaves and the most marketing it can carry.
• 3. Keep a product page standard and score every page against it before you promote it.
• 4. Make checkout predictable: delivery cost and date shown early, guest checkout, a short form and the payment methods your customers use.
• 5. Run the daily order run twice a day against a dispatch cut-off and a promised dispatch date.
• 6. Choose couriers with a scorecard and a trial, and keep a tested backup.
• 7. Process returns with reason codes, inspection grades and refunds paid the day you decide.
• 8. Answer customers within a set time per channel, from a library of messages in your own voice.
• 9. Check stock sync, apps, speed and uptime monthly, and keep policies matched to practice.
• 10. Review the numbers every Monday and audit the whole store every month.

How do you work out what one order really earns?

Start with what the customer pays for the products after discounts. Remove any tax included in the price to get net revenue, and add the shipping the customer paid. Then subtract every cost the order causes: the landed cost of the products, packing, the courier, payment fees, cash-on-delivery fees, the cost of parcels returned undelivered and the average cost of returns. What remains is the contribution per order. It must pay for marketing, salaries, rent and software before there is any profit.

The worked example below is a composite small skincare brand selling in India, with an average order of 1,150 rupees including 18% tax as a placeholder rate, product cost at 25% of net price, 45 rupees of packing, 85 rupees of courier, a 2% payment fee, 30% of orders on cash on delivery and a 4% return rate. Use the same lines with your own numbers and currency.

How do discounts and free-shipping thresholds change the numbers?

A discount lowers the price but not the product cost, packing or courier, so contribution per order falls faster than the discount. In the example above, 15% off cuts contribution per order from about 575 to about 433 rupees, so the store needs roughly a third more orders just to earn the same total. At 25% off it needs about 70% more. Work this out before you choose the offer, and afterwards compare the discount cost per extra order with your normal contribution per order.

A free-shipping threshold is a trade too: you give up shipping income on orders above it, some customers add an item to reach it, and the offer may bring extra orders. Whether it pays depends on your own order mix. Take the share of orders already above each candidate threshold from your order export, estimate how many of the rest would top up, run the numbers per 100 orders, and trial the best candidate for two weeks before you make it permanent.

What makes a product page and checkout convert?

Clarity and no surprises. A good product page has a title that says what it is and the size or variant, a first two lines that say who it is for and the main benefit, five or more honest photos including one that shows scale, how to use it, the full ingredients or materials, a size or usage guide where people choose, a one-line delivery and returns note, reviews including critical ones, and alt text on every image. Score pages against a fixed checklist and fix the weakest before you spend money sending traffic to them.

At checkout, the Baymard Institute's research is the best-known guide. Its average documented cart abandonment rate is 70.22% across 50 studies, and among the reasons shoppers in its US survey gave for abandoning during checkout, extra costs such as shipping, tax and fees came first at 40%, followed by slow delivery at 20%, distrust of the site with card details at 19% and forced account creation at 18%. Its 2024 benchmark found the average checkout has 11.3 form fields while most sites need only eight. Show costs and dates early, allow guest checkout, keep the form short and offer the payment methods your customers already use, including cash on delivery or bank-app payments where those are normal.

How should orders flow from payment to delivery?

Give every order one status at a time (new, cash on delivery to confirm, on hold, ready to pick, packed, dispatched, delivered, delivery failed, returned to origin, cancelled) and a promised dispatch date calculated from your cut-off time. Twice a day, read the exceptions: orders past their promised date, unconfirmed cash-on-delivery orders, incomplete addresses, failed deliveries. Thirty minutes before the cut-off, every paid order gets a decision: pack it now or send a delay message today.

Pack to a written standard: one order on the bench at a time, every item checked against the pick list, liquids locked and sleeved, glass wrapped and centred, and the label flat. Choose couriers with a weighted scorecard and a two-week trial, measure on-time delivery from your own orders, and keep a live backup with a written rule for when to switch. Where you sell matters too: in the United States, for example, the FTC's Mail, Internet or Telephone Order Merchandise Rule expects you to ship within the time you state, or within 30 days if you state none, and to offer the choice to cancel when you cannot.

How do you run returns and refunds well?

Decide your policy with the people who will run it, publish only what you do, and never promise less than the law gives. In the UK, for instance, online shoppers can cancel within 14 days of receiving goods and the seller must refund within 14 days of getting the item back, according to gov.uk; the EU has a similar 14-day right of withdrawal. Faulty and not-as-described goods are covered by consumer law in most countries whatever your policy says.

Then run a simple process: log each request with one reason code, decide whether low-value or broken items need to come back at all, photograph parcels before opening, grade each item (resell, repack, write off or claim from the supplier), refund the same day you decide, and look for patterns every week. Returns are a large share of online sales: the National Retail Federation and Happy Returns reported in October 2025 that US retailers expected 19.3% of online sales to be returned that year.

How fast should an online store answer customers?

Set a first-response target per channel and measure it in clock hours, because that is how customers experience waiting. Many small stores aim for a same-day email reply and replies within an hour on chat and messaging apps during service hours, but set your own target from what you can sustain and what your customers expect. Outside hours, an honest automatic reply says when you will answer.

Every reply should include the customer's name and order number, what you are doing and by when, and one next step. Send bad news before customers have to ask, especially delays. Log every conversation with a topic and read the topics weekly: the biggest topic is usually a page, policy or process you can fix. If AI drafts replies, a person reads and sends every message, no customer data goes into the tool, and customers are told when they are talking to an automated assistant.

What should you check every week and every month?

Every Monday, spend thirty minutes on five numbers (sessions, conversion rate, average order value, contribution per order and return rate), then operations (late dispatches, delivered on promise, refunds owed, first response time), then what is coming (promotions, stock risks, launches), and end with three actions, each with an owner and a date.

Once a month, make a test purchase on a phone with a payment method you rarely use, score the store area by area, review every app for cost, speed and data access, check site speed against Google's Core Web Vitals (good means Largest Contentful Paint within 2.5 seconds, Interaction to Next Paint within 200 milliseconds and Cumulative Layout Shift of 0.1 or less at the 75th percentile of visits, according to web.dev), and check that your policies still match what you do. If you want the whole system in one place, The Online Store Operations Manual from Bodhih Training includes the workbook, calculator, forms, messages and SOPs. To measure your own operating judgement, AssessAll's Small Business Operating Decision Assessment is one option, and Jobulary's guide to if-then plans helps the Monday routine stick.

Common mistakes in running an online store

Most expensive mistakes are small habits repeated every day. These are the ones worth fixing first.

• Judging the store on revenue instead of contribution per order.
• Running discounts without knowing how many extra orders they need.
• Revealing delivery cost only at the last step of checkout.
• Relying on one courier with no tested backup.
• Telling customers about delays only when they ask.
• Logging returns without reason codes, so patterns stay hidden.
• Letting refunds wait for someone to remember them.
• Installing apps for every idea and never removing them or their data.
• Using fake urgency or hidden fees, which a growing number of countries ban.
• Pasting customer data into AI tools.