With Cyber 5 just weeks away, there is still time to see your online store through your customers’ eyes.

Imagine spending weeks getting a store ready for Black Friday. The advertising is booked, the offers are approved, and the team has worked through inventory, signage and staffing. A customer walks in, finds what they want and heads to the checkout, only to be left waiting. In a physical store, we would notice. Someone would ask what was happening, call for help or open another till. Online, that customer can leave before anyone realises they needed help.

That was the thought I kept coming back to while reading Yottaa’s The 3 A.M. Problem, a guide to preparing for Cyber 5, the five-day shopping period from U.S. Thanksgiving through Cyber Monday. Drawing on millions of shopper sessions across nearly 700 retail sites during Cyber 5 in 2025, the report examines mobile shopping, website performance and traffic arriving through AI assistants. For me, it raises a familiar retail question: are we putting as much effort into helping customers complete their purchase as we put into getting them through the door?

We are only weeks away from this year’s event. Much of the promotional planning is already in place, but there is still time to experience what we are asking customers to do and address the things that might get in their way.

Across the sites Yottaa measured, sessions declined on every day of Cyber 5 compared with 2024, while conversion improved throughout the weekend. Friday’s sessions were down 12.6%, and Saturday’s were down 16.6%. To put Friday’s decline into perspective, a site with 100,000 visits the previous year would have received about 87,400 if it followed that pattern. These are visits rather than individual customers, since one person can visit more than once, but the implication is still worth considering: fewer visits mean fewer opportunities to earn a purchase.

These findings reflect Yottaa’s sample rather than every retailer, but they offer a useful challenge to how we read our own results. When traffic falls, our first instinct is often to find ways to bring more people in. That may be the right response, but I would also want to understand what happened to the customers who already arrived. Where did they hesitate? What information could they not find? Was there a point where the experience became harder than the purchase was worth?

We can spend more to attract another visit. Before we do, we should understand how well we served the last one.

Yottaa reports that mobile accounted for roughly 75% to 80% of orders across its measured sites during Cyber 5. That is about three out of every four orders, or more, coming through a phone. Yet one of the report’s most useful findings was that responsiveness to shoppers’ actions remained a weak point, even when pages loaded quickly.

Think about the pause after choosing a size, applying a filter or tapping “Add to Cart.” The customer has done their part, but the website has not clearly responded. Do they tap again? Wait? Assume the item is unavailable? A page can look ready while still leaving someone unsure whether anything is happening.

The physical-store comparison is straightforward. The doors are open, the lights are on and the shelves look good, but when a customer asks for help, nobody answers. We would not consider that store ready simply because it opened on time. We should bring the same thinking to our websites, looking beyond whether the page appears to whether customers can comfortably complete what they came to do.

The report found that 77% of AI-referred visits landed directly on a product page, compared with 60% of visits from Google. In other words, roughly three out of four visits arriving through an AI assistant began at the product itself. AI referrals were still less than 2% of total traffic in Yottaa’s sample, so this is an emerging behaviour rather than the whole business. Even so, it gives us something useful to consider.

We often give the homepage the attention we would give a flagship store window, but these shoppers may never see it. Their introduction to the retailer could be a product description, an availability message or a delivery estimate. That page needs to help them understand both the product and what buying it will involve.

Can they work out which size or model is right for them? Is availability clear? Can they find the delivery cost and return policy without hunting around? Landing on a product page does not automatically mean someone is ready to buy, but it does mean that page has to do more of the work of welcoming, informing and helping them.

The report’s title refers to an observed AI-referral peak between roughly 2 a.m. and 4 a.m. UTC, which falls in the previous evening in Canada during Cyber 5. The practical lesson is to understand when your own customers are shopping and make sure someone, or an appropriate automated response, can act when something goes wrong.

An alert is useful, but it does not help a customer complete an order unless it leads to action. In a store, telling the manager that the checkout line is growing is only the first step. Someone still has to open another till. Online teams need the same clarity about who responds, what they can change, who provides backup and how they confirm that the customer experience has recovered.

Yottaa sells performance technology, so the guide naturally makes a case for automation. My broader takeaway is about ownership. Whatever technology we use, there needs to be a clear path from recognising a problem to resolving it, especially during the weeks when so much preparation comes down to how well we serve the next customer.

With Cyber 5 only weeks away, I would make time to shop your own website from a phone, working through a real purchase from product discovery to checkout. But I would not stop there. Shop it with a few customers and watch how they find their way through it. Ask them to look for something they would actually buy, then listen to their comments as they go.

Give them room to work it out without immediately pointing them in the right direction. Where do they look first? What do they scroll past? Which words or instructions make them pause? If they ask a question, find out what they expected to see. The moment we feel tempted to explain how the website works may be the moment we have found something worth improving.

Sometimes we are simply too close to our own business. We know what the categories mean, where the delivery information lives and which button to press next. Customers bring fresh eyes. Something that feels obvious to us may be confusing to someone seeing it for the first time, and listening to that person can tell us more than another internal review.

Bring those observations back to marketing, ecommerce, technology and customer service, then agree on the changes that matter most before the event. Some may be technical. Others may be as straightforward as clearer wording, more useful product information or making delivery details easier to find. Alongside that work, walk through what happens if a key part of the shopping journey fails on Friday evening, so the team knows how to respond.

We spend a lot of time asking how to get more people into our stores. In the weeks ahead, there is real value in spending time with the people already trying to shop with us. Watch what they do, listen to what they say and give the team a chance to act on what you learn.

The customer should not have to understand how our website works. We should understand how they want to shop.

Until next week, The Retail Rewired Team