RetailTech. Meet HardTech.
Retail/FoodTech software needs breakthrough HardTech.
No matter how much software or AI is being developed and integrated, the physical reality of labor and traditional operations in a brick-and-mortar retail still acts as the primary limiting factor.
What’s missing is a hardware platform — not a software one. Better yet, an open one.
I have been in talks with many AI and software companies with fantastic ideas and models to enhance logistics with novel insights, but the barrier to entry is high. Simply put, there is no platform to tap into the business logistics in an intuitive manner, such as particular workflows and stations, to bring about anything worthwhile or long-lasting. The lack of standards makes such a task impossible. In most cases, it’s uni-directional, making it impossible to create a physical output to trigger changes in logistical operations. That’s why there is a touchscreen feedback mechanism for every software because humans are the only ones that can perform the needed labor.
When I ran a restaurant in the middle of COVID with DoorDash’s, UberEats’, and GrubHubs’ independent touchscreens, it was an absolute patch job to the current system. These apps had no clue what was genuinely going on in the business and had forced a unique behavioral change to the customer and staff by each one of them. In this case, this was a technology-first and a people-second approach. So the market’s next step was creating yet another software layer aggregating said tools that do not solve the core problem — tapping into the logistics. But the reality is that we can’t.
Even with our real-time platform our AI and UI required continued interaction with staff members for operational integrity and tranparency. You can view the four demos here:
We are highly limited in moving the ball forward without groundbreaking hardware that elevates retail. Building software for how we think about retail today will ultimately be DOA in the coming years.
The First Movers
Apple and Tesla had to create a new hardware platform for a new phone and a car, respectively. Both elevated the capabilities of a phone, and a car could be and do.
Imagine a retailer as an iPhone — it comes standard with apps to fulfill your daily workflows. While at the same time, it offers you a plethora of other apps that can tap into the logistics of your workflow to enhance or offer more significant insights. The hardware made that possible to standardize the logistical I/O every step of the way with security and versatility in mind.
What Tesla did to the automotive industry is more about software accessibility and scalability than creating the hardware itself. The hardware is the platform where all the software magic (AI) can happen, bringing in technologies that were impossible before.
Today, software companies are still working with retail, no different from companies trying to get AI working on ICE cars. So it makes more sense why Apple is making a car.
Mark my words — retail is next.
We need an open system based on a fundamentally new architecture. At a high level, we need an AWS for brick-and-mortar retail with both hardware and software automation in mind.
Internally, at Wings, we have called it Ground-to-Cloud (G2C) architecture, creating the next-generation Autonomous Sustainable Retail (ASR.) Our first B2B system to create the ground hardware automation layer sits between the internal and external (customer) logistics to remove analog limits from workflows and open up the data to anyone to use. Slowly but surely, more automation will continue to scale from either end.
In my Path to Autonomous Retail session (below), I pointed out the importance of automating workflows to increase I/O bandwidth between stations and, ultimately, the retailer itself. But more importantly, the first step in a hardware platform is to automate workflows and create access points for any app to bring in more value than the hardware chassis itself.
Realtime AI Fulfillment Logistics
With the inclusion of the hardware platform in retail, we would now have access to data to optimize further and transparently digitize the I/O of orders, fulfillment, and pickup in the cloud. Moreover, the AI layer (first or third-party) can provide realtime predictions and optimizations to facilitate operations based on active operating staff members, orders, and expected pickup methods.
Traditional (Single) Order Workflow
Traditional retail operational workflows are based on an order-by-order basis without considering customer distance from the store, external events, internal staff performance changes, method of pickup changes, or order customization complexity — creating unpredictable wait times and debilitated customer service. Moreover, in fig. 1, HR and RD times do not exist in traditional retail. Our hardware platform exposes these times for much more optimized and predictable operations.
The hardware platform can tap into more data for a multi-layer value-add intelligence across the whole operation. The internal logistics data is now transparent to the staff and customers, creating more true-to-life wait times that are dynamically changing based on staff and external events — such as how far a customer is to the store.
This path is not about removing labor. Instead, it’s about revitalizing the economy to bring in more retail entrepreneurs with a much lower barrier to entry that operates with fewer people but is more creative and fulfilling — more love.
It wasn’t long ago smartphones were limited to experienced professionals. Now anyone can. Why not retail? Business experience, optional.
That’s how I see the future and it’s not that far off.