November 24, 2025
Image Courtesy of Kroger
Has Kroger Discovered the Limits of Automated Warehouses?
In a shift to “hybrid e-commerce offerings,” Kroger announced plans to close three of its eight automated warehouses opened with Ocado while expanding partnerships with Instacart, DoorDash, and Uber Eats.
Kroger said it is making the changes to speed up delivery times, lower prices, improve store conditions and lift operating margins. The moves are expected to improve e-commerce profitability by approximately $400 million in 2026.
Kroger will take a charge of $2.6 billion against third-quarter earnings due to the closures of facilities in Pleasant Prairie, Wisconsin; Frederick, Maryland; and Groveland, Florida, in January — and its automated fulfillment network “not meeting financial expectations.” Kroger further said it is “monitoring” its remaining facilities’ performance.
With the closure of the Groveland warehouse, Kroger will be ending delivery in Florida.
In geographies seeing a higher density of demand, automated customer fulfillment will be used support capacity needs.
Ocado stated that it would receive over $250 million in compensation following the closures.
Kroger’s Partnership With Ocado Began Years Ago, But Cost Overruns Abounded
Kroger first entered its partnership with Ocado in 2018 with a goal of building 20 robotic fulfillment centers nationwide, but the openings have proved more costly than anticipated. In September, Kroger signaled a potential retreat from its investment in automated warehouses when it announced a “site-by-site” review of the fulfillment network.
Ron Sargent, Kroger’s interim chairman and CEO, said, “We are building on a strong foundation with five consecutive quarters of double-digit eCommerce sales growth and increased profitability improvements. We are taking decisive action to make shopping easier, offer faster delivery times, provide more options to our customers, and we expect to deliver profitable sales growth as a result.”
Kroger Leans on Instacart, Uber Eats To Fill the Gap
As part of the shift, Kroger expanded its Instacart relationship as its primary delivery fulfillment provider, including recently becoming one of the first retailers to offer customers access to Instacart’s AI assistant.
Kroger, in October, expanded its relationship with DoorDash to include fresh foods, household goods, and other products versus a limited selection at some Kroger locations previously.
Kroger will also form its first partnership with Uber Eats Marketplace in early 2026.
“The pandemic bump didn’t hold, and shoppers moved back to stores quickly, so the economics of running dedicated e-commerce facilities may no longer make sense,” eMarketer analyst Suzy Davidkhanian told Reuters, adding Kroger’s partnerships with the online platforms provide it more reach among consumers.
“In this environment, partnering (with delivery companies) rather than fulfilling (orders) in-house is the more efficient path forward.”
BrainTrust
Neil Saunders
Managing Director, GlobalData
Bhargav Trivedi
Nolan Wheeler
Founder and CEO, SYNQ
Recent Discussions
Discussion Questions
Do you see Kroger’s move as confirmation that third-party delivery platforms will be the driver of e-grocery in the years ahead?
Is Kroger’s hybrid formula likely the optimal model for delivery for the grocer, or do you suspect Kroger will eventually close all its robotic warehouses?
Poll
Loading …
The first issue is that the projected penetration of online grocery was overhyped during the pandemic. Indeed, on this very forum, we had ridiculous predictions that never materialized. This ultimately means the financial basis of Kroger’s investment was flawed: the volume needed to recoup the capital costs (and the Ocado model is expensive) was not there. Second, while the Ocado model helps reduce the cost to serve in terms of picking and packing, it does not reduce the cost or time to deliver – which in the US can be high because of lower population densities, especially compared to Ocado’s home market of the UK. The pivot now is to use stores – which are sunk costs and have a far better distribution for serving customers. Robotics and automation, of course, can play a role in stores.
The first issue is that the projected penetration of online grocery was overhyped during the pandemic.
As RW’s King – well , I’m at least one of the Royal Court – of online grocery negativity it seems a pity we’ve not (yet) seen the trillion dollar question actually asked: will online ever be more than a small slice of the total?
Agree with your assessment, Neil. Many retailers jumped on the online bandwagon during the pandemic without thinking that at some point it would end and a fair number of people would move back to shopping in person. That resulted in over-investing in the online business.
Just goes to show that one size doesn’t really fit anyone very well. Perishables and fresh product benefit from that human touch…while center store product lends itself better to robotic picking. Physics says shipping time and cost are directly related to distance, which makes stores a better departure point for customer delivery, But then in store picking with those large carts detracts from the shopping experience for all the customers who choose to shop in store. Keep an eye on Amazon’s new in store – automated back room robotic picking for shelf stable product, with fresh product left for human interaction. That seems a better approach then the current all or nothing methods.
Partnering with Instacart, DoorDash, and Uber Eats lets Kroger flex capacity without carrying the full operational burden. Instead of pushing automated fulfillment in every market, they’re aligning with how people already shop and receive groceries today. In that sense, a hybrid model feels like the practical path forward.
I don’t see this as proof that third-party delivery will dominate e-grocery, but it does show how hard it is to justify expensive robotics in the U.S. unless demand is extremely dense. In many developing countries, third-party services win because labor is cheaper and delivery is fast. This is a very different cost equation than Kroger’s Ocado network.
For perishables, quick turnover favors the Instacart/DoorDash model, but plenty of U.S. shoppers still prefer picking fresh items in person. That’s why a hybrid approach makes sense: use delivery partners in dense cities, while more controlled fulfillment models can still work in lower-density markets.
Whether Kroger keeps its remaining robotic sites long term will depend on the margins. For now, hybrid isn’t just safer but it’s the only approach that matches how different U.S. regions actually shop.
Kroger’s $2.6 billion lesson is about confusing infrastructure with strategy. Strategic positioning trumps operational efficiency. Kroger invested heavily in differentiated fulfillment capability while simultaneously surrendering the customer relationship to platform intermediaries. That’s not hybrid, it’s strategic incoherence. Look, Kroger built expensive, purpose-built infrastructure to fulfill orders controlled by Instacart, DoorDash, and Uber Eats. These platforms own customer data, set delivery expectations, capture margins, and determine which grocers appear first in search results. Kroger essentially spent billions becoming a high-cost wholesale/supplier to someone else’s platform. Compare this to Walmart’s approach. They’re not just investing in automation; they’re systematically building platform infrastructure to keep control of the end-to-end customer experience. Walmart understands that stores are both fulfillment nodes and customer acquisition channels. Kroger has yet to realize that distinction.
As to third-party platforms, they won’t “drive” e-grocery. They’ll intermediate it for retailers lacking scale to compete directly. And yes, Kroger will likely close all robotic warehouses, because you can’t justify capital-intensive infrastructure after conceding strategic control.
‘The problem with automated warehouses is they only work at one speed. My manual pickers are more flexible.’
That was the critique I heard when I went on my first warehouse tour in the early 1990’s.
I have watched Ocado’s story with interest – emerging as an online-only grocer in the UK – it seems that was a Trojan Horse as they morphed into a technology company.
Ocado have done well with there moves into the US and Australia – until now.
Rumours have appeared several times that Amazon may seek to acquire Ocado – but perhaps they saw a flaw – indeed Amazon’s own move into store-side automation has taken a step-back.
It would be fascinating to understand what is behind the challenges – too many products, changing too often, massive demand fluctuations, or something else?
The biggest element will be the ‘pick-from-store’ model which offers far more flexibility due to its proximity to customers, ability to operate at different picking-speeds, and that they can be developed without the need for extensive new real-estate acquisition cycles.
As long ago as 2020, McKinsey’s research advocated this approach as making far-more sense to compete with Amazon’s ‘same-day delivery’ model. So, from one perspective at least, Amazon is the reason for the limited success of automated grocery warehousing.
Kroger’s shift away from its expensive, dedicated automated warehouses to an expanded third-party delivery model (Instacart, DoorDash, Uber Eats) is a crucial move to control costs, moving from high fixed capital investment to a more flexible, variable-cost structure. While the $2.6 billion charge represents the failure of the initial Ocado strategy, this “big bath” write-off cleans the balance sheet by removing future expenses. This strategic cleanup ensures that the company’s new asset-light model is set up to successfully deliver the targeted $400 million in e-commerce profitability by 2026.
More Discussions
Are US Small Business Owners Right To Be Optimistic About Sales Growth in 2026?
A recent Comerica survey suggested that American small business owners are optimistic about what 2026 holds…Should Retailers Be Charging Fees for Mail-In Returns?
The question of whether retailers should be charging fees for mailed returns continues to be a…Who Should Lululemon’s Next CEO Be?
Lululemon is being asked to consider a specific CEO candidate. Is this the right pick?
The first issue is that the projected penetration of online grocery was overhyped during the pandemic. Indeed, on this very forum, we had ridiculous predictions that never materialized. This ultimately means the financial basis of Kroger’s investment was flawed: the volume needed to recoup the capital costs (and the Ocado model is expensive) was not there. Second, while the Ocado model helps reduce the cost to serve in terms of picking and packing, it does not reduce the cost or time to deliver – which in the US can be high because of lower population densities, especially compared to Ocado’s home market of the UK. The pivot now is to use stores – which are sunk costs and have a far better distribution for serving customers. Robotics and automation, of course, can play a role in stores.
As RW’s King – well , I’m at least one of the Royal Court – of online grocery negativity it seems a pity we’ve not (yet) seen the trillion dollar question actually asked: will online ever be more than a small slice of the total?
Agree with your assessment, Neil. Many retailers jumped on the online bandwagon during the pandemic without thinking that at some point it would end and a fair number of people would move back to shopping in person. That resulted in over-investing in the online business.
Just goes to show that one size doesn’t really fit anyone very well. Perishables and fresh product benefit from that human touch…while center store product lends itself better to robotic picking. Physics says shipping time and cost are directly related to distance, which makes stores a better departure point for customer delivery, But then in store picking with those large carts detracts from the shopping experience for all the customers who choose to shop in store. Keep an eye on Amazon’s new in store – automated back room robotic picking for shelf stable product, with fresh product left for human interaction. That seems a better approach then the current all or nothing methods.
Partnering with Instacart, DoorDash, and Uber Eats lets Kroger flex capacity without carrying the full operational burden. Instead of pushing automated fulfillment in every market, they’re aligning with how people already shop and receive groceries today. In that sense, a hybrid model feels like the practical path forward.
I don’t see this as proof that third-party delivery will dominate e-grocery, but it does show how hard it is to justify expensive robotics in the U.S. unless demand is extremely dense. In many developing countries, third-party services win because labor is cheaper and delivery is fast. This is a very different cost equation than Kroger’s Ocado network.
For perishables, quick turnover favors the Instacart/DoorDash model, but plenty of U.S. shoppers still prefer picking fresh items in person. That’s why a hybrid approach makes sense: use delivery partners in dense cities, while more controlled fulfillment models can still work in lower-density markets.
Whether Kroger keeps its remaining robotic sites long term will depend on the margins. For now, hybrid isn’t just safer but it’s the only approach that matches how different U.S. regions actually shop.
Kroger’s $2.6 billion lesson is about confusing infrastructure with strategy. Strategic positioning trumps operational efficiency. Kroger invested heavily in differentiated fulfillment capability while simultaneously surrendering the customer relationship to platform intermediaries. That’s not hybrid, it’s strategic incoherence. Look, Kroger built expensive, purpose-built infrastructure to fulfill orders controlled by Instacart, DoorDash, and Uber Eats. These platforms own customer data, set delivery expectations, capture margins, and determine which grocers appear first in search results. Kroger essentially spent billions becoming a high-cost wholesale/supplier to someone else’s platform. Compare this to Walmart’s approach. They’re not just investing in automation; they’re systematically building platform infrastructure to keep control of the end-to-end customer experience. Walmart understands that stores are both fulfillment nodes and customer acquisition channels. Kroger has yet to realize that distinction.
As to third-party platforms, they won’t “drive” e-grocery. They’ll intermediate it for retailers lacking scale to compete directly. And yes, Kroger will likely close all robotic warehouses, because you can’t justify capital-intensive infrastructure after conceding strategic control.
‘The problem with automated warehouses is they only work at one speed. My manual pickers are more flexible.’
That was the critique I heard when I went on my first warehouse tour in the early 1990’s.
I have watched Ocado’s story with interest – emerging as an online-only grocer in the UK – it seems that was a Trojan Horse as they morphed into a technology company.
Ocado have done well with there moves into the US and Australia – until now.
Rumours have appeared several times that Amazon may seek to acquire Ocado – but perhaps they saw a flaw – indeed Amazon’s own move into store-side automation has taken a step-back.
It would be fascinating to understand what is behind the challenges – too many products, changing too often, massive demand fluctuations, or something else?
The biggest element will be the ‘pick-from-store’ model which offers far more flexibility due to its proximity to customers, ability to operate at different picking-speeds, and that they can be developed without the need for extensive new real-estate acquisition cycles.
As long ago as 2020, McKinsey’s research advocated this approach as making far-more sense to compete with Amazon’s ‘same-day delivery’ model. So, from one perspective at least, Amazon is the reason for the limited success of automated grocery warehousing.
Kroger’s shift away from its expensive, dedicated automated warehouses to an expanded third-party delivery model (Instacart, DoorDash, Uber Eats) is a crucial move to control costs, moving from high fixed capital investment to a more flexible, variable-cost structure. While the $2.6 billion charge represents the failure of the initial Ocado strategy, this “big bath” write-off cleans the balance sheet by removing future expenses. This strategic cleanup ensures that the company’s new asset-light model is set up to successfully deliver the targeted $400 million in e-commerce profitability by 2026.