In August 2023, Wilko went into administration and around 400 shops closed. CDS Superstores, which owns The Range, bought the Wilko name and its own-brand products, but not the shops. Wilko returned as a website, with its products stocked inside The Range. By 2024, CDS was opening Wilko shops again in Plymouth, Exeter, Luton and St Albans.

The first decision was commercial. Wilko was a recognised name with a known product range, so bringing it back without hundreds of town-centre leases made sense. What it asked of customers was different. Shop online, travel to The Range, or go without.

That is the customer impact of retail decisions. Most business cases consider the customer in some form: demand, value, convenience, loyalty or likely response. What they do not always make explicit is the customer ask. What will people need to do differently for this decision to work? Travel further, switch channels, accept a different product, understand a new rule, or absorb more effort? And what evidence do we have that they will?

Why the customer impact of retail decisions gets missed

Customer impact is often discussed in business decisions, but it is not always tested in a practical enough way.

A business case might refer to customer demand, convenience, value, loyalty or channel preference. Those are all useful considerations. The problem is that they can stay too broad. They do not always show what the customer will actually have to do differently once the decision is implemented.

That is where the risk sits. A customer may need to travel to a different location, switch from a shop to a website, accept a different product range, follow a new returns process, understand a new pricing rule, or do more of the work themselves. Each of those asks may be reasonable. But each one needs to be made visible before the decision is signed off.

The other issue is timing. The commercial case is usually clear at the point of approval because the expected savings, margin or operational benefit have been modelled in advance. Customer impact often shows up later, through behaviour: lower conversion, more complaints, more contacts to customer service, reduced footfall, more returns, or colleagues having to explain the same change repeatedly.

By then, it can be hard to isolate the cause. Trading conditions may have changed. Competitors may have moved. Marketing activity may have shifted. That is why you should consider customer impact before the change goes live, not treat it as something to interpret afterwards.

Rather than “Will customers like this?” A more useful question to ask is: “What does this decision require customers to do, and what evidence do we have that they will do it?”

How to check the customer impact of retail decisions before sign-off

A customer impact check is a simple way to make the customer ask visible before approving a decision. It does not need to be a new governance process or a long piece of research. In most cases, it is a structured conversation using the evidence the business already has.

The point is to move beyond broad statements such as “customers want convenience” or “customers are looking for value”, and test the practical effect of the decision. What will change for customers? What will they need to do differently? Which customers will find that easier, and which may find it harder?

Five questions are useful.

1. What is changing for the customer?

Describe the decision in plain language from the customer’s side. Not “we are optimising the service model”. A counter will close earlier. A product will no longer be stocked. A service will move online. A return will now cost money. If the change cannot be described clearly from the customer’s point of view, it is not yet clear enough to assess.

2. What are we asking customers to do?

Finish this sentence and write down every answer: for this decision to work, customers will need to… Download the app. Travel further. Choose a different product. Serve themselves. Wait longer. One change inside the business often asks customers to do three or four separate things, and the list is usually longer than the people in the room expect.

3. What are we basing that on?

Take each expectation and name the evidence behind it. Sort the answers into three groups: what you know from what customers already do, what you have an indication of from research or complaints, and what you do not know yet. “We do not know” does not stop the decision. It identifies which part of it carries the risk. Insight that never reaches the people making the decision is a common failure in itself, and we covered that separately.

4. Who carries the effort or risk?

Highlight who has to work harder because of this change, and which customers are most likely to struggle or give up. A digital process is straightforward for most people and difficult for some. A different location is manageable with a car and much harder without one. Complexity removed from the business usually lands on a colleague or a customer rather than disappearing, which is the pattern behind most process fixes.

5. How will we know if we were wrong?

Choose the signals before the change goes live. Abandoned purchases, repeat calls about the same thing, customers switching products, higher returns, or colleagues explaining the same point over and over. Then decide in advance what you will do if those signals appear: clearer information, more support, a different version, or reconsidering the decision.

Used well, these questions make the customer impact of retail decisions easier to discuss before you make them. They do not remove commercial judgement. They improve it by clarifying what the business is asking customers to do and what evidence supports that assumption.

What are customers being asked to accept?

Value is not just a price. It is the price, the products customers expect to find, and the confidence that the shop will still meet the need they came in with.

Poundland is the clearest recent example. Across 2023 and 2024, sourcing for clothing and general merchandise moved up to Pepco Group level. The rationale was sound; buy across the group, use the scale, avoid running the same buying operation several times over.

Pepco later said the new ranges had underperformed, and that it had underestimated the challenge of getting Poundland customers to understand and embrace them. Like-for-like sales fell 3.6% through to the end of September 2024, then 7.3% in the quarter to the end of December. Barry Williams, who had previously run the business, returned in January 2025 to lead the turnaround.

The key point is that the decline did not level off across the period; it accelerated. Pepco also pointed to difficult UK trading conditions, weaker sales in core grocery categories, falling margins and rising costs, so the range change is not the sole cause. It is the part the company itself identified as significant.

The decision asked customers to accept different products at the same price points. That works if people come to Poundland for the price. It works less well if they come for a particular combination of things at a price they recognise.

Poundland’s response suggests the second. It moved to clearer £1, £2, and £3 pricing, added hundreds of products to the £1-and-under range, and put £1 price points on essentials, including bread and milk, under its “Home of the Pound” positioning. That is not only a pricing change. It is a change to what is behind the price.

The same question applies to Wilko. Did customers value the brand, or the shop they could walk into while in town? CDS’s chief executive answered it publicly when the stores reopened: “The public reaction to the loss of Wilko stores was undeniable.”

Can customers see what they need to do differently?

Giving customers information is not the same as giving them a clear route through the experience. The test is not only whether the information exists, but whether customers can use it when they need it.

Two decisions show the difference, and neither depends on the business hiding information.

Superdrug launched an online marketplace in November 2022 with around 300 partner brands, widening the range without buying or holding the stock. Marketplace products are labelled, and the product page says who will sell and ship the item.

The return journey is different, though. You can usually return a product supplied by Superdrug to a store. Superdrug’s own marketplace terms state that “the contract of sale is established solely between you and the Seller”, and that “Superdrug is unable to process any marketplace product returns via our warehouse, our online help team or via any of our stores”.

So the information is there, at the point of purchase. The consequence arrives weeks later, at the point of return, when someone walks into a Superdrug with the item and finds the shop cannot help. Which? has found the same pattern across Boots, Superdrug, B&Q, The Range and Wayfair, so it seems like this is how many retailer marketplaces are choosing to work. Where a customer journey crosses between two companies, somebody has to decide how the join works.

ASOS shows the same gap in a different form. In October 2024, it introduced a £3.95 charge for customers with high return rates who kept less than £40 of an order, with a further £3.95 restocking fee above an 80% return rate. The commercial problem is real: returns carry transport, processing, checking and repackaging costs, and some stock comes back too late or in a condition that makes it harder to sell.

Affected customers were told the fee applied. They could not see the return rate being used to assess them, or how close they were to it.

The charge was meant to change behaviour. But from the customer’s side, it was not clear what behaviour needed to change. Was the issue ordering too many items, returning too high a proportion of an order, using the wrong size information, or relying on returns as part of the buying process? A charge tells customers there is a cost. It does not necessarily help them understand how to avoid it.

In January 2026, fifteen months later, ASOS added a tool that showed each customer their personal return rate in the app, with warnings as they approached the thresholds. For those fifteen months, the policy could tell a customer they had crossed a line without showing them where it was.

When support depends on customers knowing to ask

A service that exists by request is not the same as a service built into the customer experience.

For example, shoes are normally sold in pairs. For a customer with a limb difference, that standard offer may not match what they need, and buying a pair to use one shoe creates an unnecessary cost.

Schuh had already been allowing single-shoe purchases through its customer experience team. In August 2024, it made the service available across all of its stores at half the price of the pair. Chief Retail Officer Neil Partington described it as part of the company’s wider inclusion strategy.

What changed for the customer is clear enough. Before, customers needed to know the option existed and contact someone to arrange it. Afterwards, they could walk into a store and buy what they needed.

This is a useful test to apply elsewhere. How many customer needs does the business meet by exception, rather than through the experience offered to everyone? Does the customer have to know who to contact, explain their circumstances, and ask for something different before the process changes?

Some of the most valuable CX improvements are not new services. They are existing workarounds made visible, accessible and easier to use.

Customer behaviour data is evidence, but it still needs interpreting

Behavioural data is the strongest evidence available about customers, and it still needs interpreting.

Boots is a useful case because it explained the customer behaviour behind a decision. In March 2023, it announced changes to the Advantage Card that took effect in May: members went from earning four points per pound to three, alongside an ongoing 10% discount on thousands of own-brand products and an expanded Price Advantage range.

Chief Marketing Officer Pete Markey said the company had seen members using points more frequently on essentials rather than saving them, and that customers wanted “on-the-spot lower prices and instant value”.

Boots was not guessing. It had evidence of how customers were using the scheme. The question was whether that behaviour meant customers wanted a different value exchange, or whether they were using points on essentials because they needed the value sooner.

Boots published more detail in 2025. Its data shows the typical Advantage Card member spends their points twice a year, and that 74% save for at least six months before redeeming them. Half use points on everyday essentials. More than a third use them for small luxuries.

The 2025 data adds useful context to the 2023 explanation. Spending points on essentials does not necessarily mean customers wanted fewer points in exchange for immediate discounts. It may mean they needed the value sooner. And the two datasets together show that members do not use the scheme in one way: a large group saves for months, a smaller group redeems frequently, and the change will have landed differently on each.

The question that follows is: who gained and who lost? Customers who shop own-brand regularly gained an immediate discount. Customers who saved points for Christmas or a larger purchase were asked to give up one kind of value for another they may not use in the same way.

Observing what customers do tells you what is happening. It does not automatically tell you what they want, and the answer is rarely the same for everyone.

The customer impact of retail decisions is not always visible at sign-off, but it usually shows up in behaviour. Customers abandon, complain, switch, ask colleagues for help, or simply stop coming back.

Key takeaways

Describe the change from the customer’s side. Not “we are optimising the service model”, but “the counter will close earlier”, “the product will no longer be stocked”, or “returns will now cost money”. If the change cannot be explained plainly from the customer’s point of view, it is not clear enough to assess.

Make the customer ask explicit. For the decision to work, customers may need to travel further, switch channels, accept a different product, understand a new rule, or do more themselves. Write those asks down before the decision is signed off.

Separate evidence from assumption. Identify what you know from customer behaviour, what you can infer from research or feedback, and what you do not yet know. “We do not know” does not automatically stop a decision, but it shows where the customer risk sits.

Identify who carries the extra effort. A digital process may be easy for most customers and difficult for some. A different location may be manageable with a car and much harder without one. Customer impact is rarely evenly distributed.

Decide how you will know if the decision is not working. Before the change goes live, choose the signals to watch: abandoned purchases, repeat contacts, customers switching products, higher returns, or colleagues explaining the same point repeatedly. Then decide what action you will take if those signals appear.