Customers will never see your audience segments or sales planning spreadsheet. They will see a price, product availability and a payment form. If any of these fail, additional traffic can increase abandoned baskets and support requests. Before Black Friday, therefore, follow the entire ordering journey as a new shopper would: on a phone, without an account, using different addresses and payment methods.
The following fifteen checks concern how your store works, rather than the discount idea itself. Record an owner, test result and remedy for each one. Faults that prevent purchases should take priority over cosmetic changes. For offer design and price communication, see our article on planning a Black Friday promotion. The purpose here is to establish that customers can buy the offer correctly, receive their order and return it if necessary.
1. Prices and offer conditions are consistent
Open an advertised product through the homepage, category, search and direct campaign link. Compare its price, promotional eligibility and end date on the product page, in the basket and in the order summary. Check excluded variants and messages for customers who are not signed in. The expected outcome is one understandable offer throughout the journey, without conditions appearing only at payment. Keep screenshots and a test order as a reference for subsequent complaints. Also appoint someone to approve published messages before traffic starts.
2. Stock availability refers to the correct variant
Choose a product with limited stock and place controlled orders as two customers, using a test environment or an agreed production scenario. Check when the system reserves an item, how long it holds the reservation and what happens after a failed payment. Repeat for a size or colour and for warehouse synchronisation. The result should confirm that availability refers to the correct variant and that abandoned transactions do not hold stock indefinitely. Separately agree what happens if synchronisation is delayed, so support knows when to pause sales.
3. Codes and discounts combine according to the rules
Build a small combination table covering promotional prices, newsletter codes, loyalty discounts, free delivery and excluded products. Test permitted and rejected combinations, including after changing quantities or removing an item from the basket. The expected result is the correct final total and a clear explanation when a code is rejected. Disable unintended combinations before launch. Do not rely solely on the written terms when the basket mechanism behaves differently. Keep examples that let your team reproduce a fault quickly when a customer reports it.
4. Delivery charges appear before the final step
Build baskets below and above the free delivery threshold, including heavy products and different collection methods. Check whether applying a discount changes the threshold calculation according to your chosen rule. Customers should know the delivery cost before spending time completing the full form. Any progress bar showing the amount still required must use the same value as the order summary. Show surcharges when the delivery method is selected and verify the final total in the order confirmation. Repeat the check after applying a promotional code.
5. Delivery promises match the address and timing
Test addresses in your main sales areas and a location with restricted service. Check courier and collection point availability, along with the dispatch message for orders placed after the cutoff time. Distinguish the dispatch date from estimated arrival. The expected outcome is a promise your warehouse staffing and carrier schedule can fulfil. Decide who changes the message when the order queue grows. A fixed next-day dispatch claim should not remain visible when the team cannot honour it, including over the weekend.
6. Mobile purchasing works from arrival to payment
Complete the journey on a real phone, using a typical browser and a mobile connection. Check the menu, variant selection, add-to-basket button, form keyboard and return from the payment app. Watch for promotional banners or consent windows covering the continue button. The expected result is a purchase without zooming or guessing where to tap. Also test arrival directly on a product page through an advertising link: this session may bypass important information and offer explanations on the homepage that your regular customers already know.
7. Guest checkout is genuinely available
Start a fresh session without saved details. Select guest checkout and make sure the form does not require a password or an extra registration confirmation. Also test an email address associated with an existing account and automatic form completion. The expected result is ordering without mandatory sign-in wherever the store promises that option. Ask someone outside the team to try without instructions. Their hesitation at an unclear button may reveal a problem that employees who know the store no longer notice in everyday work.
8. Failed payments allow a safe retry
In an agreed test mode, reproduce a declined payment, an interrupted redirect and a delayed provider confirmation. Check the order status and the option to retry payment from a message or account area. Customers should understand whether money was taken and what to do next. Retrying must not silently create another payable order. Also establish when support may manually change the status, so goods are not dispatched simply because the customer returned to the website before the system received the payment provider’s confirmation for that transaction.
9. Confirmations do not create duplicate orders
Test a double click on the purchase button, refreshing the confirmation page and repeated delivery of a payment notification in a controlled scenario. Compare order numbers, stock reservations and customer messages. The expected outcome is one valid order with an unambiguous status. The confirmation should include products, the total, delivery method and store contact details. Ensure support sees the same reference as the customer and can identify a real duplicate quickly, before the warehouse packs two parcels or the system requests a second dispatch of the goods.
10. Transaction tracking does not count a purchase twice
Place a labelled test order and compare the store dashboard, analytics tool and advertising reports. Verify the transaction identifier, value, currency and handling of events sent from both browser and server. Refreshing the thank-you page should not increase purchase counts. Define whether reporting covers placed, paid or accepted orders. The expected outcome is an understandable difference between reports and a way to reconcile data, rather than forcing every dashboard to match despite different definitions and different times for attributing sales to advertising campaigns.
11. Measurement respects consent choices
Open the store in a fresh session and inspect behaviour after acceptance, rejection and changing an earlier choice. The technical team should confirm which mechanisms run in each scenario and whether the configuration matches the adopted consent policy. Displaying a banner alone does not demonstrate correct operation. The expected outcome is respect for user decisions and documented data limitations. Do not fix reporting discrepancies by firing every tag regardless of the customer’s choice. Record the test scenarios so a store update does not reintroduce a configuration fault you previously removed.
12. Performance has been tested under agreed conditions
Agree a safe testing scope with the administrator, ideally in an environment that reflects production. Assess response times for categories, product pages and baskets, but also message queues and order synchronisation. Do not run unapproved load tests against the public store or payment services. The expected result is identified limitations and a decision about who responds when they are exceeded. Check monitoring, backups and hosting contact availability. A fast homepage alone does not prove that the store can handle simultaneous payments and stock updates effectively.
13. Forms explain errors and remain accessible
Navigate the basket using a keyboard and check visible focus, button descriptions and form labels. Enter an invalid postcode, leave out a required field and select an unavailable variant. Messages should identify the specific problem and explain the correction while preserving information already entered. The expected outcome is independent checkout, including for people using assistive technology. Do not communicate through colour alone. If an error appears outside the visible phone screen, direct the user’s attention clearly to the place where action is required.
14. Support understands help and returns procedures
Ask an adviser to resolve three sample requests: a missing confirmation, an incorrect size and a partial return from a promotional order. Check contact visibility, working hours and access to payment and dispatch statuses. The expected outcome is an answer based on the same rules the customer saw. Prepare short escalation instructions and a realistic response-time message. If calculating a refund requires someone else’s decision, identify that person beforehand so the adviser does not promise an amount without checking the specific offer conditions and order contents.
15. Ending the promotion and reverting changes are prepared
Rehearse ending the campaign on a copy or in an agreed scenario. Check prices, codes, banners, advertising links and automated messages after expiry. The expected result is a consistent return to the regular offer while preserving the conditions of orders already accepted. Identify who can stop sales of an affected product and restore the previous configuration. Instructions should specify the action sequence and how to verify the result. Prepare them before the weekend, when technical contractors may have different availability from an ordinary working day.
Close the review with one register: scenario, result, correction owner and release decision. Not every fault must stop the entire campaign, but every critical issue needs a solution or an explicit restriction of scope. Connect the preparation order to the action plan through Cyber Monday, and assess the impact of faults using the sales and margin calculation.
If you need support reviewing the purchase journey, MB Digital Marketing can help identify obstacles and prioritise changes. Let us discuss e-commerce marketing together with your store’s readiness to handle orders. One realistic basket, several payment methods and a list of problems to resolve before increasing traffic can provide a practical starting point.