Infrastructure, the Gateway — It Starts with the Customer
Shopper analytics, customer service, loss prevention, shelf restocking and same-day fulfillment are all in the realm of real-time retail requirements today. Retailers have a variety of analytics that can indicate what needs to happen to increase in-store sales. That is an urgent kind of analytic, not an end-of-the-month review. For example, the number of customers in the store now, their location (so you can direct sales associates to assist them), or an alert that current inventory is running low, so you can restock the shelf now.
We have the software to do all that. The challenge is implementing the infrastructure in the store to collect the data and use it immediately to manage the merchandise and the employees on the floor.
At one end is a full portfolio of cameras, smart shelves, scanners and hand-held readers, sales associate tablets (for inventory lookups, pricing and in-aisle checkout), RFID and sensors all over the store, magic mirrors and kiosks, facial recognition, and edge servers with all the software in each store. That kind of infrastructure takes a lot of power, too.1 Then these systems have to be installed, calibrated, and maintained. They make sense for large general merchants2 or extremely high-end brands3 or concept stores like Amazon Go.4 But for other retailers, even with thousands of stores, the cost and personnel required to take care of all that may not make sense.
Yet, all retailers want to increase sales and reduce shrinkage. An area critical to saving the store is to provide what the web cannot — service, the ability to try things on, and getting skilled advice about what to buy. Shoes, underwear, electronics, prepared foods, and furniture are examples of categories that require a high level of customer service. So, knowing that a customer is in the store is critical, because one of those important metrics — fitting room conversion — may be missed without that service. (This metric is based on the stats observing that once a customer tries something on, they are more than 60% more likely to purchase. This metric is also true for tasting stations and other ways that fully engage the customer.) For that conversion to happen, your salesperson needs to be “on the job.” For chains with hundreds or thousands of stores, but with small footprints, installing all that infrastructure is a bit much. Fortunately, there are solutions for all sizes of footprints.
![]() |
Kepler Analytics Retail People Counter Sensor |
At recent retail demonstrations, we saw many in-store shopper analytics alternatives being highlighted. At one end of the spectrum is a solution like Kepler Analytics. They have retailers that absolutely fit the profile of what we are talking about above — that high need for customer care, for example at Hanes or Bras N Things.5 These retailers have the experts to make sure you get the products that will work for you.
Another interesting retailer that Kepler supports is Assembly Label from Australia. They have a unique philosophy that is environmentally responsible and focused on the customer experience. In these types of enterprises, data is critical to ensure that their values are always achieved for each and every touchpoint.
As for the infrastructure, Kepler’s Store Operation Toolkit is a deceptively simple-looking device6 backed up by the latest in Machine Learning. The store-wide device collects the “ping” from mobile phones. This signal does not swipe data off your phone. The goal here is just to track the in-store analytics, not collect your personal data and track you. Many consumers (the general public and governments) are becoming alarmed about the lack of privacy and, therefore, security, that our mobile and web revolution has wrought upon us. It is not only the lighter infrastructure, but the protection of privacy that makes Kepler’s approach an important alternative to the facial recognition trend we are seeing.7
At the other end of the spectrum are Sensormatic, Honeywell8 and even some of the RFID firms like Impinj that integrate to partner software platforms. Yes, there clearly is a value proposition here. In larger stores, shrinkage is generally quite a bit more than in those with a smaller footprint. (In larger stores, as well, it is more difficult to notice that customer who may be wandering around looking for a fitting room.) Those tech firms have been very successful, not just with the infrastructure, but in providing the data platform for applications.
Retailers have to determine what type of solution is a fit for them, including the cost of the personnel resources to maintain the infrastructure, not just the upfront start-up implementation costs. Increasing the profit per square foot and product margin is key here, so determining the blend of technologies to achieve that objective should be the focus in infrastructure decisions.9
I Robot. You Jane. Into the Warehouse
Logistics is also getting a big make-over, too. The warehouse has been a source of innovation in automation for decades. High racks, sortation conveyors, RFID, pick to light, and so on have been a mainstay of the large, high-velocity facility. But now, with “one each picking,” same-day delivery, and more delicate and perishable items being bought online, as well as fulfillment centers that support multiple brands, new solutions are popping up in the warehouse.
Roving robots are arriving in more and more facilities.10 As we all know, Amazon has an astronomical 200,000 robotics drives, already. (Amazon’s competitors use terms like “collaborative robots” to describe their robots, which we found curious.) Although the manufacturers of these types of robots call them autonomous, the fact is that they do need to work with — collaborate with — humans a lot more than is advertised. Although there might be an ROI on “time to fulfillment,” the human cost may be the bigger issue here.
We find it oh so cute to hear Pepper and Temi talk to us. But Robots, as yet, have challenges with even basic physical interactions with people, like carrying a cup of coffee without spilling it, or more to the point, working safely with people in the warehouse. A recent report by researchers Beth Gutelius, PhD., and Nik Theodore, PhD, at the University of Illinois, focused on the changing worker landscape in the fulfillment warehouse, leading to some rather stark conclusions about the relentless production pace, safety issues, and thus, higher employee turnover.
Conclusions: Infrastructure — the Real Cost
Unlike in retail, where there is the perception that the technology will allow sales associates to spend more time with customers, which should increase job and customer satisfaction, in the warehouse, robotics will likely de-skill workers, leading to less job satisfaction and lower pay.
In fact, the dynamics of work are changing radically due to technology on the floor and in the office. Executives, as they contemplate investments in automation, need to think more broadly about the overall returns — not just shareholder value, but stakeholder value, which includes their workforce and customers.
Managers and supervisors, who are on the front lines of these changes, need to be the ones who assess the blending of technology and people in a way that supports these broader values. The C Suite will not be able to understand the details, the ergonomics, and the conditions in the facilities. That is the manager’s job to understand and explain.
Tech and their service providers also need to be part of the solution. We are so used to selling products and services with those big ROIs — faster, cheaper — that sometimes we miss broader values. As we are all humans, we do need to bring our humanity into the equation to support the consumer’s need for privacy, worker safety, job satisfaction, sustaining the environment, and delighting the customer.
__________________________________
1 Installing these may also mean a significant “rewiring” job. — Return to article text above
2 think Neiman Marcus, Saks, etc. — — Return to article text above
3 think Channel, Tiffany — Return to article text above
4 check out the photo essay by Bill McBeath at the Amazon Go store in Seattle. — Return to article text above
5 a retail chain of women’s undergarments in Australia, Asia, and Africa — Return to article text above
6 Kepler says about their people counter that it is a “compact & wireless device invented, designed, & built in Australia by our engineers. The CBS works as a people counter & provides business intelligence grade data via a real-time web dashboard — ” — Return to article text above
7 Note this statement from Kepler Analytics on their website: “With all this in mind, something critically important to us is maintaining absolute protection over shopper privacy. We live in a world full of invasive tracking systems — from security cameras to apps on our phones. While we can’t stop what others are doing, we can set a positive example and reject the notion that privacy is something we should ever compromise on. We are proud sponsors of the Privacy by Design project headed by Dr. André Gygax from the University of Melbourne and will continue to do our part to support consumer privacy protection.” — Return to article text above
8 Who have diverse portfolio of technologies, with Sensormatic being able to deliver an entire portfolio of infrastructure plus an AI-powered retail analytics platform — Return to article text above
9 Firms like Checkpoint and Sensormatic have had decades in loss prevention, so their hardware, support, and advice for reducing shrink and improving inventory is invaluable. — Return to article text above
10 Read the Robot Report, 2020. — Return to article text above
To view other articles from this issue of the brief, click here.