The Retail Note - Amazon Fresh: on the turn

Key Messages

• Amazon Fresh to close its UK business
• Five of the 19 stores to be rebranded as Whole Foods
• Format launched to major fanfare in March 2021
• ‘Just walk out’ technology a questionable USP
• Inability to scale-up a key reason for failure
• Amazon too late to a highly competitive market
• C-store retailing far more complex than credited
• Brand and technology only take a retailer so far
• Amazon remains a minnow in UK grocery market
• Losses continue to mount at Whole Foods
• Amazon yet to crack transition to physical retail.

Amazon (not so) Fresh. Amazon Go(ne). Two of the slightly crueller headlines I could have used, although the one I went for doesn’t exactly score highly on the kindness scale. Any apparent smugness on my part stems less from Amazon’s failure in this instance, more on general despair with the retail analyst community that I represent.

The barriers to entry to becoming a retail analyst (a retail expert even) these days are depressingly low. We are all consumers, we all go shopping, in a digital age anyone can put finger to keyboard or face to camera and express an opinion on retail. Retail analysts and so-called “experts” are now ten a penny. But those that actually understand the very fabric, dynamics, economics and finances of retailing remain few and far between.

Many of the armchair analysts are little more than Amazon cheerleaders. And they were certainly out in force when, after months/years of speculation, Amazon finally launched Amazon Fresh in the UK back in March 2021. They waxed lyrical at how good the format was, wondered at the deployment of ‘frictionless’ check-out free technology and boldly predicted that it would completely revolutionise store-based retailing. I quote some verbatim: “this move will have the major UK grocers quaking in their boots”. “Amazon always plays to win in everything it does. If it plays in UK grocery, it will win”.

For now, Amazon hasn’t won. UK grocery retailing remains un-revolutionised. Tesco et al are probably sleeping fairly soundly. And those armchair analysts are now looking very silly.

The facts

Before my take on the whys and wherefores, the facts. The advent of Amazon Fresh was the subject of fevered anticipation for a very long time before it actually materialised. The first store opened on Ealing Broadway in 2021.

Rather than a massive offensive hub and platform for the online non-food business that many armchair analysts predicted, Amazon was at pains to stress that Fresh was absolutely a food-first convenience store (c-store) format. And rightly so. The ambition was for it to be a business that could stand on its own two feet, rather than something that merely piggy-backed on Amazon’s wider e-commerce ecosystem.

As a c-store format, it offered both branded and own-brand lines, with a product mix spanning both ambient and fresh, including the higher margin food-to-go and ready-meal options that tend to make c-stores more financially viable. But the main selling point was the ‘just walk out’ technology, whereby cameras picked up what shoppers selected and charged them to their app, obviating the need for them to go to a check-out. That was the revolutionary part.

Slow but steady expansion ensued and the portfolio crept up to around 20 sites. Then the first signs started to appear that the format was perhaps not living up to expectation. In July 2023, the business announced a handful of store closures, including the original store on Ealing Broadway that had seen people queuing round the block only two years earlier. And, if not publicly, a moratorium was called on new expansion.

In between, a significant U-turn in that checkouts were installed in the remaining stores, the format effectively giving up on its technological USP. And then everything went deathly quiet until a couple of weeks ago, when Amazon announced that it would pull the plug on the residual 19 stores. Not a complete withdrawal from UK grocery by any means, as five of the Amazon Fresh’s will be converted into Whole Foods stores. Plus, obviously Amazon still plays in online grocery (to a degree) and has various partnerships with the likes of Morrisons, Co-op, Iceland and Gopuff.

Why did Amazon Fresh fail?

Not because consumers shunned “big brother” technology, as the tabloid press would have us believe. But for two, more basic and highly interlinked, reasons. 1. C-store retailing is a very tough and competitive market with a wide margin of error. 2. Brand and technology will only take you so far.

The barriers to entry for c-store retailing are deceptively low – anyone can open a small shop selling groceries. But to actually make a success of it and make money, so many retailing disciplines have to be spot on. Above all, c-store retailing operates on an entirely different business model from supermarkets, with far fewer economies of scale to fall back on. The classic mistake is to think that c-store retailing is just a scaled-down version of mainstream grocery retailing.

The cost base of c-store retailing is much higher than larger formats, if not on an absolute / total basis, certainly on £/sq ft metrics. Securing the right location at the right rent is absolutely key and there is huge scope for getting this wrong. With a much smaller footprint to play with, the product mix has to be absolutely right for that location. By definition, c-stores have to stock everyday essentials (milk, confectionary, alcohol), but these are low margin and in some cases, loss-leaders. The art is to counterbalance these with products that are higher margin – fresh, food-to-go, ready meals etc – but with these come a whole host of other complications, such as effective supply chain, storage, shelf-life and wastage. A whole host of considerations and delicate balances to strike.

Amazon entered the c-store bear pit largely blind and woefully late. It had very limited grocery experience, certainly relative to the competition, and was having to effectively learn on the job. In all fairness, it did make a decent fist of this, the format itself wasn’t bad, nor was the assortment and product mix shabby. But not bad and not shabby is not good enough in a market that is fiercely competitive.

The fact that Amazon was late to the c-store game was more a factor in its ultimate demise than any shortcomings in the format itself. Tesco launched Express way back in 1994 and largely had the market to itself until Sainsbury’s got its act together with Local. Tesco alone has more than 2,000 Express stores, plus ca. 700 One Stops, plus however many thousand Londis, Budgens and Premier stores that came its way through the acquisition of Bookers. All the major UK grocery incumbents play in the c-store arena to some degree and have been aggressively scouring the market for opportunities for years. And continue to do so.

By the time Amazon joined the party, the pickings for new sites were relatively slim. With most of the best opportunities long gone, Amazon did have some extra leverage with some landlords on the basis of its brand name. But it was never going to scale up to the degree that it needed to in the face of the market that was already there. A key failing of Amazon Fresh was that it was in a catch-22 situation from the get-go. It needed to scale up to achieve the requisite economies of scale to be profitable. But the more stores it opened, the greater the risk of opening unprofitable ones.

The only way of breaking out of this catch-22 situation would have been through acquisition of a competitor – a course of action Amazon did not choose to take.

The limitations of brand and technology

Amazon’s brand is unquestionably one of the strongest on the globe, obviously transcending far more than retail. Many of the armchair analysts mistake this as making Amazon an infallible, omnipotent force, against which other retailers are powerless to compete. Amazon Fresh serves as a reminder that this is not true by any means.

Amazon is a formidable retail brand in non-food. But in grocery? In base terms, would you opt for a Tesco or M&S branded sandwich or an Amazon one? A Costa or Starbucks coffee or an Amazon one? A McDonalds burger of an Amazon one? Some may take the Amazon option in each case, but they would not be the majority.

Undeniably clever as it was, the ‘just walk out’ technology was always more of a gimmick than a game-changer. At the time, I think I dismissed it as a solution to a problem that didn’t exist, a monicker that others have since adopted. Largely all c-stores these days have self-checkouts and the process is hardly arduous. In terms of cost savings, Amazon Fresh still needed staff to replenish the shelves and, ironically, to help customers use the app. They may as well have been serving customers behind a till than standing dutifully by the entrance.

As for the revolutionary technology, the bottom line was it was extremely expensive, to the point of undermining a store’s viability. Tesco and Sainsbury’s both trialed it, but didn’t adopt it on this basis. One grocery property director once explained to me that it would only ever work in stores with a certain configuration and largely doubled general fit-out costs. A hit-and-miss market where the right store and location is paramount should not be comprised and led by technology requirements.

A fact that is often lost on the armchair analysts, technology can only be an enabler in retailing, it cannot be an end unto itself. The current groundswell towards AI needs to be framed in the same context. AI will no doubt be a force for good in retail in that it will vastly improve efficiencies. But it will not be a substitute for core retailing disciplines. In other words, AI will not make bad retailers good.

What next?

The armchair analysts and Amazon cheerleaders have either gone very quiet or suddenly gone into denial, with the benefit of hindsight saying why Amazon Fresh was never going to work. But some are sticking to their guns, saying that Amazon will come again in UK grocery.

And who’s to say it won’t? But for the time being, it remains an absolute minnow in the UK grocery market. The online business continues apace, but the UK online grocery market isn’t as high growth as many predicted it would be, with a penetration rate stagnating around the 8-9% market and general profitability unproven.

And, of course, Amazon still owns Whole Foods. But eight years on from acquisition (2017), the business has shown very few signs of progress, certainly in the UK. After considerable chopping and changing, there are just six Whole Foods stores, excluding the proposed five Amazon Fresh conversions. The business is heavily loss-making and has been for years. In 2024, the business reported a pre-tax loss of ca. £20m, on top of a loss of £23m the previous year. Apparently, total losses have accumulated to more than £200m since the business launched in the UK in 2004. Sales also went backwards in 2024.

Amazon to gain a strong position in the UK grocery market through acquisition? If it was going to do so, the question is why it didn’t when it had the chance, on the relative cheap? It may still happen (Ocado?) – but then so too may a total withdrawal from the UK physical retail market.

One final thought. Online vs physical retail debates are now so hackneyed to the point of being largely irrelevant. Retail is multi-channel, omni-channel or whatever the armchair analysts want to call it. Most store-based retailers have made the transition to become multi-channel operators, one way or another. The transition has been painful and the journey is never-ending, but it has been made nonetheless. Conversely, it is a transition that Amazon has yet to fully make.

Multi-channel. Amazon – supposedly omnipotent, but struggling to achieve something that its store-based peers have successfully managed?