Can ‘Amazon Now’ succeed where Fresh stores failed?

Joining the likes of Uber Eats, Just Eat and Deliveroo, Amazon has launched its own rapid grocery service in the UK. Amazon Now delivers thousands of products such as milk, eggs, over the counter medicine and nappies within 30 minutes.

It is “the fastest delivery experience in Amazon’s European history” says Amazon EU quick commerce program lead Elisa Michelin Salomon.

Initially available in select postcodes in Southwark in south east London, the dark store, micro-fulfilment model is just the beginning of a wider rollout, “laying the foundations for what’s coming next in the UK and the EU”.

But why Amazon Now, now? Can Amazon make it succeed after a long line of grocery flops? And how worried should competitors be?

On the face of it, Amazon’s rapid play appears badly timed. It’s long missed the opportunity to swallow up the slew of 15-minute delivery, dark store operators that were struggling when the sector consolidated in 2021, following an explosive, pandemic-driven ascent.

And it’s a little late to be leveraging a network of stores to fulfil orders – as per Tesco Whoosh, Sainsbury’s Chop Chop and Morrisons Now – following the closure of its 19 Amazon Fresh outlets at the end of last year.

“Those locations would’ve been perfect micro-fulfilment hubs,” says independent e-commerce consultant Matt Anderson. “Close to customers, ready to go. But they were bleeding money. High rent, retail staffing, poor footfall.”

Many Fresh stores were in prime retail locations, like Kensington High Street and outside Liverpool Street Station.

Yet there is method in the apparent madness. Its first dark store, named QLD 1, is understood to be situated under railway arches. Just as close to shoppers, but much cheaper to rent. “Amazon realised they were paying retail overheads to run what should’ve been dark stores,” Anderson adds.

By waiting, it’s also allowed plucky newbies to do the hard work – and spend the marketing money – to convince consumers they need groceries to be delivered immediately.

“It was sensible for Amazon to take a wait-and-see approach,” says Natalie Berg, Amazon expert and founder of consultancy NBK Retail. “This is a wildly capital-intensive model and we’ve seen failure after failure. Five years ago, no one really knew where we’d land, but it’s clear that quick commerce is here to stay.”

The mainstream nature of the rapid delivery market is proven by rivals. According to a September white paper from Just Eat, two in three UK consumers say they are likely to order groceries for on-demand delivery in the next 12 months, and 72% say fulfilment in under an hour helps them ‘take back control’ of their lives.

Anderson agrees Amazon has allowed others to “burn billions proving the quick commerce model”. The likes of Getir and Gorillas “all failed building demand, infrastructure and unit economics simultaneously” as “Amazon just waited”.

“[Amazon] watched them educate customers, watched them fail, then walked in with something none of them had: millions of existing UK customers,” he adds. “No customer acquisition cost to amortise across £12 baskets. No need to convince people fast grocery delivery is worth paying for. The heavy lifting’s done.”

The wait-and-see strategy seems to have worked for Tesco. It launched a pilot with rapid delivery company Gorillas in 2021 before rolling out its own Whoosh service to 1,000 Express stores in 2023.

Amazon, meanwhile, had a 11.5% stake in Deliveroo until its by US rival DoorDash in October. It also counts rapid player Gopuff as one of the handful of grocers with a storefront on Amazon.co.uk.

Grocery credentials

There are other benefits, too. Amazon also doesn’t need to convince consumers of its grocery credentials. One in 10 already buy groceries to some extent from Amazon, according to October research from Levercliff. It also found that nearly half of those already doing so were switching more of their spend to the retailer.

And unlike established aggregator apps, which have offered swift delivery of groceries from partner supermarkets for several years, Amazon is going its own way with a dark store model. Only Deliveroo has used the model locally, at its handful of Hop sites.

The advantages of using dark stores – more accurate picking and inventory management, as well as speed – would make a ‘no substitutions guarantee’ more than feasible.

Berg believes Amazon is going “back to the drawing board” to figure out its point of difference, following a series of failed experiments.

“Amazon’s USP in grocery has always been hard to define – not the cheapest, not the most premium, various partnerships, and a muddled UX,” says Berg.

For her, this latest move is about Amazon is leveraging its reputation for convenience. “Amazon is about reducing the shopper’s cognitive load – this is what Amazon can do that others can’t. This is what Amazon is famous for,” she says.

What’s more, Amazon has already made Now work in India and the UAE.

The service launched as a pilot in Bengaluru in December 2024, later expanding to Delhi and Mumbai. The company has now opened more than 100 micro-fulfilment centres across the three cities, with plans to “add hundreds more” in other urban hubs as “order volumes continue to grow at 25% month on month” it says.

Now is available “across all major locations” in UAE since launching in January last year.

“But these are both markets where quick commerce has been much more disruptive and where labour is cheaper than the UK,” cautions e-commerce consultant Viv Craske.

It makes “onlookers wonder what Amazon knows about unit economics of quick commerce, despite the failure of dark-store, quick-commerce models in countries with higher wage bills”, he adds.

Craske argues Amazon Now offers a poor user experience compared with rivals such as Deliveroo and Whoosh.

“Amazon Now’s Online shopping experience is nothing new, and without innovating in online UX, it’s unlikely to drive demand for more efficient, larger baskets. And so I don’t see how Now’s profitability will fare better than previous dark store quick commerce models.”

However, Craske adds, it’s quite possible Amazon “knows the rapid delivery model is inefficient but expects to make up the losses in increases demand for retail media advertising”.

As Anderson puts it: “Thin grocery margins don’t matter when you’re making it back through Prime subscriptions, sponsored products, and the wider basket that follows.

“The structural advantages are real: demand density from day one – no cold start problem, or expensive cost per acquisition, as the Prime customers are already there. Infrastructure is already built – local nodes, routing tech and courier networks. But here’s what they still need to prove: enough profitable missions per micro-catchment to cover courier costs at scale.”

“The real game Amazon is playing isn’t about competing on speed. It’s about resetting what ‘normal’ feels like.”