Walmart’s Alphabot Robot Is The Beautiful First Letter Of A New Retail Language | Forbes (a disagreement)
Forbes published this online article: Walmart's Alphabot Robot Is The Beautiful First Letter Of A New Retail Language
I disagreed with the article and had a comment exchange with the author:
What WalMart is doing isn't the 'end of the world' but it is a very bad decision that will NOT work in the long term. It will waste resources and be a distraction that delays what really needs to be done.
Building a 20,000 sq ft extension to an existing store and stuffing it with equipment and inventory is the worst possible implementation of 'OmniChannel'. The goal should be to have one platform that can serve instore sales, delivery and outside pickup. In a business where margins are single digits, two side by side stores duplicating each other is not going to beat the merchant who gets it right. And getting it right is a single integrated store front.
Walmart's investment in this approach is doomed and the experience gained will not serve as a base for the real solution. Walmart needs to recognize Moore's Law. Every couple of years costs halve and technical capability doubles. Walmart needs to invest in the right approach in a very modest number of stores until they can make it work. And they can be first if they want.
I can easily visualize several 'picking' solutions for existing style stores. The simplest to understand is the robot occupying human space picking items from shelves just like any other shopper. More likely solutions are operating robots from overhead accessing a 2 or 3 feet space corridor behind existing product shelves. Robots would not interact with humans but use the same store and inventory. And any 'picking' solution also solves the restocking of shelves.
There are NO moon shot technical hurdles to make this work. Just the application and integration of known technologies.
In the 5 to 10 years it takes to get it right, Walmart might have to have manual processes in many stores to stay competitive and to keep improving customer facing apps. But at today's level, human picking labor will be far less expensive than capex for buildings and equipment and associated operating costs.
==========
This post documents the comment exchange on the blogger Chris Walton's site Omnitalk.blog
=========
Comments exchange from the comments blog
Chris Walton: "...I disagree with you, however, on your points about there being technologies that could do this within current operations. If companies are not willing to go all in on the redesign, stand alone warehouses at least get past the biggest hurdle – inventory accuracy. Automation solutions are difficult to deploy inside stores because inventory is constantly mixed back of house and front of house. Inventory has to be conceived of and kept separate for anything to work..."
My response:
Inventory accuracy won’t be an issue beyond stockouts caused by theft and other small transient conditions. POS and delivery will accurately enough show consumption and standard inventory systems are handling inputs well today. Front / back of house splits are relatively easy to handle when robots do the shelf restock and existing systems are modestly modified to treat front/back as separate ‘locations’. And robot shelf auditors will keep transient conditions from causing unexpected stock depletion. No ‘rocket science’ here.
I disagreed with the article and had a comment exchange with the author:
What WalMart is doing isn't the 'end of the world' but it is a very bad decision that will NOT work in the long term. It will waste resources and be a distraction that delays what really needs to be done.
Building a 20,000 sq ft extension to an existing store and stuffing it with equipment and inventory is the worst possible implementation of 'OmniChannel'. The goal should be to have one platform that can serve instore sales, delivery and outside pickup. In a business where margins are single digits, two side by side stores duplicating each other is not going to beat the merchant who gets it right. And getting it right is a single integrated store front.
Walmart's investment in this approach is doomed and the experience gained will not serve as a base for the real solution. Walmart needs to recognize Moore's Law. Every couple of years costs halve and technical capability doubles. Walmart needs to invest in the right approach in a very modest number of stores until they can make it work. And they can be first if they want.
I can easily visualize several 'picking' solutions for existing style stores. The simplest to understand is the robot occupying human space picking items from shelves just like any other shopper. More likely solutions are operating robots from overhead accessing a 2 or 3 feet space corridor behind existing product shelves. Robots would not interact with humans but use the same store and inventory. And any 'picking' solution also solves the restocking of shelves.
There are NO moon shot technical hurdles to make this work. Just the application and integration of known technologies.
In the 5 to 10 years it takes to get it right, Walmart might have to have manual processes in many stores to stay competitive and to keep improving customer facing apps. But at today's level, human picking labor will be far less expensive than capex for buildings and equipment and associated operating costs.
==========
This post documents the comment exchange on the blogger Chris Walton's site Omnitalk.blog
=========
Comments exchange from the comments blog
Chris Walton: "...I disagree with you, however, on your points about there being technologies that could do this within current operations. If companies are not willing to go all in on the redesign, stand alone warehouses at least get past the biggest hurdle – inventory accuracy. Automation solutions are difficult to deploy inside stores because inventory is constantly mixed back of house and front of house. Inventory has to be conceived of and kept separate for anything to work..."
My response:
Inventory accuracy won’t be an issue beyond stockouts caused by theft and other small transient conditions. POS and delivery will accurately enough show consumption and standard inventory systems are handling inputs well today. Front / back of house splits are relatively easy to handle when robots do the shelf restock and existing systems are modestly modified to treat front/back as separate ‘locations’. And robot shelf auditors will keep transient conditions from causing unexpected stock depletion. No ‘rocket science’ here.
Two completely separate structures requires more inventory than one. Two buildings are more expensive than one. The winner where profits are measured in single digits is going to be the dense single location that services in store customers, pickup and delivery. And WalMart should only work to this solution, every other path makes it harder to compete with Amazon.
The transition will be a long 15 to 20 year cycle but the successful local site level path will be clear in the mid 2020s.
Comments
Post a Comment