By September, most beauty and fashion brands have already started talking about Black Friday, Christmas gifting and the promotions they want to run, but the conversation is normally led by the campaign calendar rather than the condition of the website that is expected to convert all of that demand. Offers get agreed, media budgets get increased and email plans get filled, while the basic customer journey is left until somebody notices a problem during the busiest week of the year.

That order is backwards. Peak trading does not make a weak website stronger, it puts more pressure on every part of it, including navigation, search, stock messaging, product information, mobile speed, checkout and customer service. A small issue that loses a handful of orders in a quiet week can become a very expensive problem once traffic doubles and paid media is at its most competitive.

I would start with the journeys that are likely to receive the most traffic, not a general redesign or a long wish list. Identify the products, collections and landing pages that will sit behind the largest campaigns, then test the full route from advert or email through to checkout on mobile. The important question is whether a customer can understand the offer, find the right product, make the necessary choice and complete the order without needing to stop and work anything out.

Peak exposes catalogue problems quickly. Products with inconsistent names become harder to search for, weak filters make large gifting collections difficult to use, out-of-stock variants waste the first positions in a collection and bundles break when one component becomes unavailable. These are not design details. They affect whether the traffic you have paid for can turn into revenue.

The same applies to promotional setup. Every offer needs to be tested against existing discount codes, loyalty rewards, free delivery thresholds, bundles and automatic discounts because customers will find combinations the trading team did not consider. The terms should be understandable before checkout, and the basket should explain exactly what has been applied without making the customer calculate it themselves.

Operational information also needs to be visible where the customer makes the decision. Delivery cut-offs, returns periods, gift options and stock availability should not be buried in a policy page, particularly once customers are buying against a deadline. If a product cannot arrive before Christmas, say so clearly. Hiding the issue may protect conversion for a few hours, but it creates cancellations, complaints and support work immediately afterwards.

Once the main journeys have been checked, create a short peak backlog and give every issue an owner and deadline. Separate what must be fixed before traffic increases from what would merely be nice to improve. A broken discount rule, unclear delivery message or poor mobile collection deserves attention before a new homepage animation.

The final part is deciding what will be monitored once activity starts. Agree the daily view now, including sessions, conversion, revenue, contribution margin, stock, discount cost, payment failures, returns signals and customer service themes. If nobody knows what good looks like or who can make a decision, the team will spend peak reacting to screenshots and isolated Slack messages.

The campaign calendar matters, but it should sit on top of a customer journey that has already been tested. More traffic will not repair the site for you, and November is a very expensive time to find that out.