I was reading an interview with B&Q’s retail director in the Retail Gazette about the company’s upgrades to its showrooms. Beautiful bathroom and kitchen displays, scent diffusers that emit a gentle coffee aroma, and aspirational spaces designed to elevate the customer experience. The kind of spaces you might want to photograph.

They’ve also added separate entrances for their TradePoint customers and dedicated routes for local builders, tradespeople and DIYers in the middle of a project.

On the surface, it’s a practical move about convenience. But I think it reveals something deeper about how customers change as they progress through projects.

Someone covered in door dust, midway through stripping a wall or building furniture, walking into a perfectly polished showroom sends the message: “This place isn’t for you right now.”

The separate entrance strategy acknowledges this. Customers move from needing the showroom in the dreaming stage to needing something more straightforward in the doing stage.

Same retailer, same goal of helping customers. But different intelligence for different moments.

That’s what this episode is about. The difference between online shopping vs in-store shopping, and how the two channels work together for the same customer at different points.

We’ve spent years trying to make online shopping and in-store shopping feel the same. Each channel has developed its own way of reading the customer, its own intelligence. And most of the time, these two sides don’t know what the other one is seeing. They don’t line up, and they don’t learn from each other.

I’ve found six patterns by comparing physical stores and online shopping, and understanding these insights might help you consider how multiple channels can work more smoothly together.

Pattern One: Nudges vs Feeling

The architecture of online shopping is built on behavioural economics. Every webpage you encounter has been designed, tested, and refined to guide your decision-making process through psychological triggers. “Only 2 left in stock” creates urgency through scarcity. “Bestseller” badges leverage social proof. Price comparisons anchor your perception of value before you’ve even considered whether you need the item.

These techniques work. Research consistently shows that behavioural economics tactics can increase average order values by up to 30%.

B&Q built 10 different AI-powered recommendation algorithms for their website, including frequently bought together suggestions, personalised product recommendations, and predictive surfacing of items based on browsing patterns. More than 10% of their e-commerce sales now come directly from these algorithmic nudges. The system tracks what you’re viewing, what’s in your basket, and what you abandoned last week, and calculates the next thing you’re statistically likely to need before you’ve consciously thought of it.

Now consider what B&Q did when they examined their physical stores with the same analytical lens they applied to their website. Their response was to create separate TradePoint entrances that bypass the showroom entirely, serving local builders and DIY customers. They repositioned help desks at eye level instead of behind counters. They recognised that the same customer needs different environmental intelligence at different stages of the same project.

The contrast reveals something fundamental about how online shopping vs in-store shopping operate on entirely different kinds of intelligence.

Digital nudges behaviour through data patterns and psychological triggers. It’s relentless, precise, always learning. The system knows your browsing time, your basket contents from last Wednesday, which products typically convert when shown together, and what time of day you’re most likely to complete a purchase.

Physical space holds feeling through environmental design and human judgment. It’s slower, more contextual, and harder to scale. The welcome doesn’t come from an algorithm. It comes from whether someone covered in plaster feels like they belong there. Whether the aisle layout makes sense when you’re carrying a sheet of plywood. Whether staff make eye contact or look straight past you.

Both approaches work. Both serve legitimate purposes. But they require entirely different kinds of investment and attention. The question we should ask ourselves is: are we designing purely for conversion outcomes, or are we considering how the experience feels to customers at each stage of their journey?

That distinction matters more than most organisations realise.

Pattern Two: Pattern vs Presence

This pattern reveals itself in what retailers call “personalisation”, a single word that describes two fundamentally different capabilities.

John Lewis partnered with Mapp Fashion to build a personalisation system that generates up to 100 million outfit combinations every night. The algorithm doesn’t just match products together; it reads demographic signals, purchase history, browsing patterns, and contextual data to predict what specific customers might want to see. A younger shopper gets nudged towards Mango. Someone browsing premium categories sees more LK Bennett. The system tailors itself based on occasion, price sensitivity, style preferences, and dozens of other variables processed simultaneously. It operates at a scale no human team could ever match.

Then there’s Lush.

Lush gives its store staff the freedom to hand over a product, to just give it away, no payment required, if someone looks like they’re having a difficult day, celebrating something special, or if a colleague has simply built a genuine connection with them. No manager sign-off required. No approval process to follow. No conversion targets to hit. It’s a decision made in the moment by someone reading the room, responding to presence rather than data patterns.

This is what psychologists call “thin slicing”. The human ability to make surprisingly accurate judgments about someone’s emotional state, confidence level, and immediate needs from minimal environmental cues.

These snap judgments aren’t perfect, and they’re certainly subject to bias. But they contain genuine intelligence that algorithmic personalisation consistently misses: real-time contextual awareness of how someone is feeling right now, in this moment, in this environment.

In the customer experience work I do, I see this all the time. There are people on shop floors who are remarkably skilled at reading customers; noticing who needs space, who’s looking for connection, who’s overwhelmed by choice, who’s confident and just needs quick information. They often don’t recognise what they’re doing as a specific, valuable skill. But it is.

Lush has maintained this random acts of kindness policy for years. It’s woven into how they hire, train, and evaluate performance. Customer testimonials consistently cite these moments as primary reasons for returning. The lifetime value impact of one £8 bath bomb given to someone having a rough Tuesday can’t be captured in quarterly conversion metrics, but it shapes long-term loyalty in ways that algorithms struggle to replicate.

The contrast becomes even more apparent when you look at how each approach handles product discovery.

Both forms of personalisation work. Both serve legitimate commercial purposes. The algorithmic approach enables efficiency and reach that human observation can’t scale. The presence-based approach captures nuance and contextual intelligence that patterns can’t access.

The question is whether you know which kind of intelligence matters most at different stages of the customer journey, and whether insights from one are informing how you think about the other.

Pattern Three: Intent vs Hesitation

Digital retail has become extraordinarily sophisticated at reading customer behaviour. Every click leaves a data trail. Every page visit, hover duration, scroll depth, and return to the same product gets logged and analysed. Shopping basket behaviour, time spent comparing options, exit pages, it’s all captured with precision that would have seemed impossible twenty years ago.

But what about what customers don’t do? That’s often where the real decision-making happens, and it’s the space where online and in-store shopping reveal completely different kinds of intelligence.

75% of UK online shoppers report inconsistent sizing across brands. For someone buying shoes or clothing online, this creates a specific kind of paralysis. They hesitate, second-guess, then implement a self-protective strategy: order three sizes, return two. Sometimes all three get returned because none fit quite right.

ASOS and Next see this behaviour constantly. It’s become one of the biggest cost pressures in UK fashion e-commerce. I’m guilty of this myself, especially when buying for specific occasions. I don’t even think about it in the moment. I find what I think I like and add it to my basket. Then I might go back to it later and reassess. I might buy multiple sizes, check which ones fit, and return the rest to the store. It’s become an accepted part of buying clothes online rather than a sign of a poor customer experience.

The economics behind this behaviour are staggering. UK retailers lose an estimated £34.4 billion annually to basket abandonment alone, which is up £2.9 billion from the previous year. This figure doesn’t even include completed purchases followed by returns.

Online systems can track this hesitation with remarkable precision. They know which products are repeatedly added to baskets but never purchased. They can identify customers who’ve viewed a product across multiple sessions over the course of weeks. They measure the exact moment someone abandons checkout and which friction point triggered it. Retailers use this intelligence to send reminder emails, offer discounts, show “customers also bought” recommendations, and more to convert hesitation into purchase.

But the data can’t tell you why that hesitation exists in the first place for that person in that moment.

In a physical store, hesitation looks entirely different. It’s someone picking up a shoe, turning it over, checking the price tag, then putting it back on the shelf. It’s a customer glancing around for staff assistance and not seeing anyone available. It’s time spent in an aisle that ultimately never converts into a sale, leaving no data trail beyond foot-traffic patterns that store managers rarely examine in detail.

Most physical stores still operate primarily on end-of-day sales figures. They know precisely what’s been purchased. They have detailed inventory management showing what’s moved off shelves. But they have almost no systematic visibility into what was considered and rejected. What questions never got asked? What concerns never got voiced. What might have sold if someone had been there to notice the moment?

Some retailers have started recognising this intelligence gap and building systems to bridge it.

Currys found that 58% of people shopping for expensive tech want to speak with someone before committing to a purchase. A £900 laptop or a £1,500 television is a significant purchase, and customers have specific questions that product descriptions don’t always address. How does this perform for my actual use case? Will this work with the equipment I already own? What happens if something goes wrong?

So Currys built Shop Live, a 24/7 video chat service connecting online customers directly with store-based product experts in real time. Not chatbots following scripted responses, but actual human beings who can read the customer’s uncertainty, pick up on questions that haven’t quite been articulated yet, and respond with context-aware expertise.

The results revealed something important: 57% of shoppers who received this kind of assistance reported higher satisfaction levels, and return rates dropped measurably. The hesitation wasn’t a friction point to be optimised away through better UX design. It was a signal that someone needed human intelligence to complete their decision-making process.

This reveals a fundamental challenge in omnichannel retail. Intent looks clear in the data (high engagement, repeated visits, extended browsing time), but hesitation often remains invisible until it manifests as abandonment. Physical stores have the opposite problem. They can see hesitation in real time but often lack the systems or attention to respond to it effectively.

The question worth sitting with: who in your organisation is paid to notice what’s not being said? What hesitation signals are visible in one channel that could inform how you design the other?

Pattern Four: Friction as Failure vs Friction as Care

A UK shopper will wait an average of five minutes and 54 seconds in a physical queue before abandoning their purchase. Regional variations reveal interesting tolerance patterns: Liverpool customers top the patience table at six minutes 47 seconds, while Plymouth shoppers drop off earlier at four minutes 55 seconds.

Online, the tolerance threshold collapses dramatically. If a checkout page takes more than three seconds to load, 57% of shoppers abandon their baskets entirely. 43% report they won’t return to that retailer after experiencing a poor online checkout process.

The friction-reduction philosophy makes intuitive sense. Remove obstacles, smooth the path to purchase, and make buying as effortless as possible. Amazon’s one-click ordering became the gold standard precisely because it eliminated nearly every friction point between impulse and completion.

But the same customer who expects instant checkout online will willingly queue for six minutes in a physical store for exactly the same product. Same person, same purchase, but completely different tolerance for friction depending on whether they’re shopping online or in-store.

This suggests friction itself isn’t the problem. Context determines whether friction feels like failure or like care.

Some retailers have started recognising that friction can serve purposes beyond creating obstacles. They’re making it deliberate.

The IKEA effect is well-documented in behavioural economics research. When people invest effort in creating something (even just assembling flat-pack furniture), they value the end result significantly more than identical items they didn’t build themselves. The effort doesn’t just create the product; it creates emotional attachment.

IKEA’s entire business model leans into this insight. Self-assembly isn’t a cost-saving measure passed on to customers as value. It’s a fundamental part of how customers engage with the products. The effort of building your own bookshelf creates psychological ownership before it’s even standing in your home.

Apple understands friction-as-care from a different angle. The way an iPhone box opens is meticulously designed. That resistance of the lid, the precise arrangement of components underneath, the careful choreography of unwrapping – it’s all intentional. It makes you slow down. It signals that this object was crafted with attention, that it deserves your attention in return.

This is friction serving as communication. The extra seconds it takes to unbox an iPhone aren’t efficiency failures. They’re creating a moment of anticipation and ceremony that positions the product as valuable before you’ve even turned it on.

Compare this to Domino’s pizza tracker, arguably one of the most successful digital friction-management tools in UK retail. The tracker doesn’t eliminate waiting time. Your pizza still takes 25-35 minutes from order to delivery. But it transforms how that friction feels. The waiting becomes active rather than passive. You’re watching your order being prepared, seeing it go into the oven, following the driver’s progress.

The friction (waiting) remains the same, but the experience of it has changed completely.

This reveals a more nuanced understanding of friction in retail: not all friction is equal, and not all friction should be eliminated. The question isn’t “how do we remove all obstacles?” It’s “which friction points communicate care, and which ones signal indifference?”

Online shopping has been relentlessly optimised for frictionless efficiency, which works brilliantly for transactional purchases, where speed and convenience drive value. But for purchases that carry emotional weight, have symbolic meaning, or require genuine expertise, that same frictionless approach can feel disposable rather than premium.

Physical retail hasn’t necessarily figured out how to use friction intentionally. Too often, queues and wait times represent operational failures rather than strategic choices. But physical spaces do offer something digital channels struggle to replicate: friction that feels human rather than systemic.

So, where in your business does friction feel like care? And where does it feel like customers are just waiting?

Pattern Five: Path vs Memory

Digital retail is optimised for efficiency, getting someone from discovery to checkout as quickly and smoothly as possible. Every step in that customer journey gets measured, tested, and refined. Conversion funnel analysis breaks the experience into discrete stages: landing, browsing, consideration, basket, checkout, and confirmation. Each step represents an opportunity to reduce friction, eliminate drop-off, and move customers closer to purchase.

This approach works brilliantly as a path. It’s clean, logical, and measurable. You can see precisely where customers abandon, identify bottlenecks, implement fixes, and track improvement.

But the purchase path and a buying memory aren’t the same thing.

Daniel Kahneman, Nobel laureate and behavioural economics pioneer, discovered what he called the peak-end rule. We don’t remember experiences as an average of every moment we live through. We remember two specific points: the peak (the most intense moment, whether positive or negative) and the end (how it concluded). Everything else fades into background noise.

This has profound implications for how retail experiences get remembered and whether customers return.

IKEA applies this principle. Walking through an IKEA store can be a source of friction for some: long pathways, deliberate routing that prevents shortcuts, the physical effort of navigating multiple floors, carrying purchases, and loading your car. By traditional customer journey logic, this should tank satisfaction scores.

But at the exit, there’s an ice cream stand and the IKEA food store. It’s a small thing, relatively low cost to operate. But it rewrites how the whole trip feels when you think back on it later. The peak (finding that perfect bookshelf) and the end (leaving with ice cream or Swedish meatballs) become what you remember. The exhausting walk through the homewares sections you didn’t need fades into background detail.

Hamleys on Regent Street operates on similar peak-end principles but executes them differently. Seven floors, live demonstrations, interactive displays, and theatrical presentations. The entire environment is designed to create memorable peaks. Moments of delight that children (and adults) talk about afterwards. A visit to Hamleys becomes an event, not just a shopping trip. Parents bring children back specifically for the experience, not just to purchase toys they could order online more conveniently.

Lush’s Liverpool flagship store created something remarkable during Christmas 2023. They transformed an entire floor into an immersive maze of seasonal scents, lighting installations, soundscapes, and textures to touch. The space became something you couldn’t replicate through online shopping. A peak experience designed to drive footfall during the year’s most competitive retail window. More importantly, it gave people something to talk about, share on social media, and remember when thinking about where to shop for gifts.

In digital retail, by contrast, the end is usually an order confirmation page. A functional screen. A place to access your receipt and tracking number. It serves its operational purpose, but it creates no memory worth retelling. The peak, if one exists at all, probably occurred while you were browsing and found something you wanted. The purchase conclusion itself holds no emotional weight.

Research on the peak-end rule in retail has found that its effect remains powerful even when customers can logically assess that the ending was disproportionately brief relative to the overall experience. A 90-minute shop doesn’t get averaged out by a 2-minute checkout. That checkout carries disproportionate weight in how the entire experience is remembered and whether customers choose to return.

The contrast between physical and digital retail on this dimension is stark. Physical stores accidentally create peaks all the time: a particularly helpful associate, an unexpected discovery, a pleasant interaction, or even well-designed spatial experiences. These moments emerge organically from the environmental context of being in a space with other people.

Digital retail, optimised for efficient paths, tends to eliminate peaks by design. Frictionless checkout means nothing slows you down enough to create a memorable moment. Algorithmic recommendations surface products logically but rarely surprise you with genuine discovery. The entire experience flows smoothly from intention to completion without creating emotional events worth remembering.

This suggests a fundamentally different question for omnichannel retail strategy: are you designing for completion or for memory?

Both matter, obviously. Paths need to function. But if customers complete transactions efficiently but form no memorable connection to the experience, are they likely to choose you again when dozens of alternatives offer equally smooth paths?

Here’s a question for you: what stories do customers retell about their experiences with you, rather than just complete them?

Pattern Six: Ignite vs Validate

Product trends used to build gradually. A new item would get press coverage, perhaps celebrity endorsement, and maybe feature in magazine editorials. It would work its way through the fashion system over months before appearing in stores, usually timed to seasonal retail peaks such as Christmas or summer holidays. The pathway from discovery to availability was measured, predictable, and controlled by retailers and brands.

That model has collapsed.

TikTok Shop became the UK’s fourth-largest beauty retailer in 2024, with beauty sales growing 60% year-on-year through 2025. A product can go from completely unknown to sold out in major retailers within days, sometimes hours. 51% of UK consumers report purchasing something specifically because an influencer recommended it. The velocity of trend ignition has fundamentally changed, and traditional retail hasn’t fully adjusted to the implications.

This represents more than just faster trend cycles. It reflects a fundamental shift in where product discovery happens and how purchase decisions get validated.

Digital platforms (TikTok, Instagram, YouTube) have become where trends ignite. The algorithm surfaces content to millions of viewers simultaneously. One viral video can create demand that didn’t exist 48 hours earlier. Trends ignite online, but customers still want physical validation before fully committing.

When Apple launches an iPhone, customers have already found content online, such as technical reviews, unboxing videos, feature comparisons, and influencer opinions. They’ve seen the product from every conceivable angle on screens. Yet queues still form outside Apple stores on launch day because people want to hold it, feel the weight, see the actual screen quality under real lighting, and judge the size against their hand. The online content creates desire; the physical validation closes the decision.

Superdrug recognised this dynamic early and responded strategically. They became the first UK retailer to install TikTok creator booths in stores, initially at their London Lakeside and Rushden Lakes locations. Customers can walk in, pick up products that have gone viral online, and immediately film their own content right there in the shop. The retail space is transformed into a content studio, a place to validate what the algorithm has shown them while simultaneously participating in the content cycle.

Boots went further than simply stocking viral products. They trained over 4,000 pharmacists across their UK network on dermatological skincare; the same conditions and ingredients customers were learning about through TikTok and Instagram content. Skin barrier health, ceramides, acne treatments, and rosacea management. When someone walks in asking about something they’ve seen online, they don’t just get pointed toward a product display. They get an actual conversation with someone who understands the science behind their question.

This represents strategic validation infrastructure. Social platforms create awareness and desire at scale that retailers can’t control or replicate. But retailers can design validation mechanisms that convert social discovery into committed purchases.

The Works implemented something simpler but equally effective. They created dedicated BookTok sections. Separate shelving for books that had gone viral on TikTok. This physical curation acknowledged where discovery happened while making it easier for customers to find socially validated products in-store. The strategy helped them grow their book market share even as the wider publishing sector struggled with declining high-street sales.

But this ignition/validation dynamic creates significant operational challenges.

44% of UK audiences report losing trust in influencers when content feels scripted or overly produced. This creates a paradox: the authenticity that makes social content effective for ignition also generates scepticism that requires validation before purchase commitment. Customers simultaneously trust and distrust the same content.

65% of retailers report out-of-stocks when products go viral because demand spikes faster than supply chains can keep up. Researchers have started calling this the “buzz gap”. It’s the operational disconnect between the velocity of algorithmic trends and retail logistics. A product that took six months to work through traditional marketing channels now needs to be available everywhere within 48 hours of going viral, but inventory systems aren’t designed for this speed.

The trust dynamics reveal something fundamental about this pattern.

Customers trust social platforms for discovery because content feels authentic, peer-driven, and algorithmically surfaced rather than brand-controlled. But this same authenticity creates hesitation to purchase. Is this genuinely good, or just algorithmically popular because it photographs well? Will it work for my specific needs, or only for the exact people I saw demonstrating it? Are the positive experiences real, or has the algorithm selectively shown me content from the small percentage of users it worked for while hiding the disappointed majority?

Physical retail and established e-commerce sites offer different trust signals: return policies, professional customer service, verified reviews with fraud detection, expert curation, institutional reputation, and the ability to examine products in person. These aren’t exciting trust signals. They’re functional, institutional, and slower. But they close sales that social discovery alone frequently can’t convert.

The implications for retail strategy are clear: you can’t control where trends ignite anymore. That’s happening on platforms with algorithms optimised for content virality, not purchase conversion. But you can design validation mechanisms that close the gap between social discovery and committed spending.

The question for you is where in your customer journey does digital ignite desire, and where do customers seek validation before actually purchasing? Are those validation mechanisms designed as intentionally as your discovery strategies? And when products go viral on platforms you don’t control, does your infrastructure respond fast enough to convert momentum into sales before the trend moves on?

So what role is your physical space playing in the trends your customers are already following online?

Six Patterns, Two Kinds of Intelligence

So there you have it. Six patterns across two very different channels.

• Nudges versus feeling – guidance towards a path versus how you feel in a space
• Pattern versus presence – personalising by knowing data or spotting signals face-to-face
• Intent versus hesitation – how customers go from “this is what I need” to “not sure about it now”, and that’s different whether you’re online or in store
• Friction as failure or care – where inefficiency online is seen as failure, whereas inefficiency in store might be part of the experience
• Path versus memory – are we creating just the path to purchase, or are we helping our customers make memories?
• Ignite versus validate – trends being ignited in one space but validated in another

I think these are patterns worth noticing.

Two kinds of intelligence. One is built on data, one’s built on presence. But they’re working in the same organisation, and most of the time they don’t know what the other one is seeing.

That gap between online and in-store shopping shows up in ways that cost money, trust, and opportunity.

Some of these patterns reveal things that need fixing. Some might reveal strengths you didn’t know you had. The trick is knowing which is which.

But maybe the real question is how to design experiences that let each channel do what it does really well, and ensure a seamless handover between them.

What if you started getting curious about what the other channel already knows?