![\"Retail](\"https:\/\/dmsretail.com\/RetailNews\/wp-content\/uploads\/2022\/05\/RETAIL-ONLINE-TRAINING-728-X-90.png\")
<\/a><\/p>\n<\/p>\n
When Amazon chief executive Andy Jassy told analysts in October 2025 that AI would accelerate the shift from physical retail to ecommerce, essentially discounting any viable future for stores, he was making a prediction that suited Amazon\u2019s view of the world.<\/p>\n
Retail, in that version of the future, becomes faster, more automated, more predictive and less dependent on the concept of needing to travel to a physical location. The customer asks, the machine finds, the platform fulfils. The shop floor becomes a relic of an earlier, messier age.<\/p>\n
It\u2019s perhaps a compelling argument, given the ever-rising market share of online. However, it\u2019s also fundamentally flawed.<\/p>\n
Yes, there\u2019s no denying that AI is changing physical retail, and quickly. It\u2019s currently doing a stellar job of exposing tired stores, weak service, poor stock accuracy and retail models that depend on customers tolerating friction. But that\u2019s not the same as saying that it will ultimately make stores irrelevant. In many cases, it will do the opposite.<\/p>\n
The real threat is not to bricks-and-mortar itself, but boring bricks-and-mortar. The store that offers little more than shelves, queues and indifferent service will continue to struggle in an AI-shaped market \u2013 just take TG Jones as a prime example. If a customer can compare prices, read reviews, check availability and receive a product without leaving home, a poor store has very little left to defend. Habit won\u2019t be enough, location won\u2019t be enough, even range won\u2019t be enough.<\/p>\n
But the best stores have always offered something harder to automate. They give customers confidence via vital human connection, they edit choice with physical comparison, they create immediacy, and they offer increasingly important advice, reassurance, human judgement and a reason to spend time with a brand in the real world.<\/p>\n
A recent ICSC and McKinsey report argues that as AI takes on more of the search and comparison process, stores are likely to become places for fulfilment, returns, product validation, immediate access and differentiated experiences. The same research found that 68 per cent of consumers had used at least one AI-enabled shopping tool in the previous three months, with 62 per cent using it to compare brands, models, prices or reviews.<\/p>\n
In other words, the customer entering the store may already know more than they used to. That raises the bar for physical retail. It doesn\u2019t remove the need for it. The store visit becomes less about basic information and more about certainty. Does this fit? Does it feel right? Is it worth the money? Can I trust this recommendation? Can someone help me choose without overwhelming me? Those aren\u2019t questions that disappear because AI exists.<\/p>\n
This is where the industry needs to be careful. Too much of the retail AI conversation still slips into a replacement mindset. Fewer people, fewer interactions, fewer costs. More automation presented as progress.<\/p>\n
There will be areas where automation makes obvious sense. AI can improve forecasting, stock allocation, staff scheduling, returns processing and fulfilment. It can help retailers understand local demand faster and remove some of the operational drag that makes stores expensive and inconsistent.<\/p>\n
The commercial case is strong; research cited by McKinsey and ICSC suggests that AI, automation and advanced analytics could increase retailers\u2019 gross operating profit by up to 20 percentage points. AI-based demand forecasting could reduce stock-outs by 20 per cent and cut inventory costs by up to 10 per cent. Real-time in-store data could increase conversion by 15 per cent.<\/p>\n
If AI can give associates better product knowledge, help managers plan labour more intelligently and make stock more visible across channels, it frees people to focus on the parts of retail where they make the greatest difference. Advice. Service. Problem-solving. Selling with confidence. Reading the customer in front of them.<\/p>\n
That is why the smartest retailers won\u2019t use AI to gut the store, but to make the experience more efficient and effective for customers.<\/p>\n
Tecovas offers a useful example. The western wear retailer uses AI in its Boot Runner programme so store associates can request specific boots and have them brought to the sales floor in around 85 seconds. The technology does the fetching. The human stays with the customer during a purchase that can feel unfamiliar for a first-time buyer. As Tecovas\u2019 chief technology officer Kevin Harwood put it, the brand has used AI to create a better human connection in store.<\/p>\n
That should be the test for retail AI. Does it protect the customer relationship, or does it weaken it? The answer will separate good strategies from lazy ones.<\/p>\n
The recent Andon Market experiment in San Francisco shows why full automation is a much messier prospect than the theory suggests. Andon Labs handed an AI agent called Luna a budget, a corporate card and the task of running a physical store. Luna could hire contractors, source products and make certain logistical decisions. It also exposed ethical and operational problems, including failing to consistently disclose that it was an AI in hiring conversations. At the time of reporting, the store was not profitable, with estimated monthly operating costs of roughly 14,300 dollars against revenue of 6,000 to 8,000 dollars.<\/p>\n
What does this prove? That a store is more than a collection of technical decisions held together by inventory, labour and margin. It\u2019s things that remain intaginble for an analytical technology. Mood, timing, trust, presentation, local knowledge, human behaviour and thousands of small judgements made across a trading day. The best store managers know when a display is not working before the data catches up. The best associates know when to step forward and when to leave a customer alone. The best retailers know that efficiency is valuable, but only if it improves the experience rather than stripping it bare.<\/p>\n
AI can support those judgements. It can\u2019t, and likely will never be able to, replace the need for them.<\/p>\n
This is the strongest counterpoint to Jassy\u2019s argument. AI may accelerate ecommerce, but it will also force physical retail to become more deliberate. Stores will need to know what they are for. Some will be built around convenience, with fast collection, returns, availability and operational ease. Others will be built around discovery, with curation, service, theatre and expertise. Many will need to combine both.<\/p>\n
What they cannot be, is vague.<\/p>\n
For years, too many stores have survived by being present rather than purposeful. That era is ending. If AI gives consumers more power before they arrive, the store must give them something better when they do.<\/p>\n
That could be a faster trip, a better answer or a more confident purchase. It could be a product they can try, touch or take away immediately, perhaps a member of staff who knows enough to challenge the algorithm rather than simply repeat it. It could be a space that makes the brand feel more tangible, more trustworthy and more useful.<\/p>\n
AI will become part of almost every serious retail operation. It will shape how retailers forecast demand, allocate stock, plan labour, personalise service and connect physical shops to digital channels. But in bricks-and-mortar, its highest value will not be in replacing the human encounter but materially making that encounter better.<\/p>\n
The weak store will be exposed. The lazy store will be punished. The store that treats AI as a way to cut its way to mediocrity will find customers have better options elsewhere.<\/p>\n
But the store that uses AI to become sharper, faster, more informed and more human has a very different future. And it\u2019s this, that Andy Jassy simply doesn\u2019t understand.<\/span><\/p>\n