Harry Haysom/Getty Images When Amazon opened its cashier-less concept grocery store to the public in Seattle on January 22, the company’s stock rose 2.5% – slightly more than the 2.4% increase after Amazon announced that it was buying Whole Foods this past June. Does Amazon Go signal that smartphone apps and virtual carts will replace checkouts in grocery stores? A low-cost, automated store is certainly an experiment worth watching. But retailers shouldn’t rush to rip out their registers just yet. Retail is littered with promising technologies introduced with great fanfare that didn’t become mainstream because they didn’t sufficiently benefit either the retailer or the customer. One reason is that new tools often don’t save the retailer enough or generate sufficient new revenue to cover their cost. Another is that too many customers simply don’t like them. To become the new normal, a technology has to make it beyond the early adopters – typically 10 to 15% of retail shoppers – and appeal to the less tech-savvy majority. Amazon Go may face challenges in both respects: It is unclear its eventual savings will justify the investment, or whether customers will find the digital surveillance it entails to be a tolerable price to pay for an improved shopping experience. Our survey of 1,500 British consumers revealed that 19% would consider an app that tracked them an invasion of privacy.  So before adopting any technology, retailers must answer two questions: How does this benefit us? And will consumers like it? Let’s look at how retailers have — or have failed to —  answer these questions. Radio-Frequency Identification Starting in the mid-1990s, RFID was expected to transform inventory control by using electromagnetic fields to automatically identify and track tags attached to objects. It also promised to obviate the need for manual barcode scanning at checkouts. But the tags never took off, because their cost was never offset by sufficient benefits. Consumers were satisfied with bar scanning for most products, and retailers found better alternative processes to track their inventory. At the British retailer John Lewis, for example, store staff today can scan an item of clothing in a store to find out what sizes are available on the shelf, in the back room, and in a warehouse for online delivery, and then send a customer all the details by email. In grocery, where checking out is the most cumbersome part of shopping, the cost of putting RFID tags on inexpensive items made the technology a non-starter. When every penny of cost matters on a case of shipped product, spending more for simple RFID tags or for intelligent tags that track temperature and other data points just does not make economic sense. Self-Checkout Self-checkouts have proved more successful. But more than 10 years since their introduction, most consumers choose them only if lines at conventional checkouts are a big hassle. Retailers have nonetheless embraced them where space is a constraint or labor costs are high. Self-checkouts have become commonplace in convenience stores in the United Kingdom where customers are in a hurry and other parts of Europe, such as the Netherlands, where floor space is limited. But fewer retailers and consumers in the United States consider self-checkouts worthwhile, even in dense urban areas such as New York City, because labor costs and the average value of sales per square foot are usually lower and the lines are just not that bad. Moreover, the limits on what customers and machines can do themselves reduce the effectiveness of self-checkout. Staff are still needed to check some customers’ bags to reduce theft and to OK their IDs for buying, say, a bottle of wine. Customers find this mildly inconvenient, while retailers have to employ people, reducing the labor savings they might have expected. Electronic Shelf Labels Electronic shelf labels allow store managers to change price displays more easily or to have them update automatically. They are slowly catching on, but it’s taken a decade. The early tags were expensive and often difficult for customers to read, and they were only adopted in markets such as France where store staff are paid relatively well, making the labor savings compelling. But now that upfront installation cost of the tags is declining, more retailers are experimenting. The displays are improving, allowing retailers to show more product information and promotional offers. Today’s labels also have the potential for making real-time price changes that respond to things like surges in supply or expiration dates. Mobile Payments There is a lot of excitement around the use of mobile phone apps, such as Apple Pay, to make payments. Many retailers are considering their own payment apps in part to avoid credit card fees and encourage customers to pre-pay for purchases before they even reach the store.  But relatively few consumers use these apps. While many are comfortable with them, most still find their chip, or contactless, credit and debit cards just as easy, fast, and secure. To encourage the use of apps, some retailers, including Starbucks, are starting to offer inducements such as extra loyalty points. The stock market was excited about Amazon Go because it signaled a promising new technology by a digital retail powerhouse that is aggressively moving into physical stores. If Amazon is right, investors are betting, the model could transform traditional retail. But it won’t happen unless the benefits to retailer and customer decisively outweigh the cost of the technology That may not be a major concern for a deep-pocketed company like Amazon, which is perpetually trying new things, learning from mistakes, and then moving on to the next potential big thing. Amazon has already benefited from the industry buzz around its concept store, which could very well set customer expectations for the in-store experience of the future. But other retailers will most likely find the initial costs too high and benefits too uncertain to immediately follow in Amazon’s footsteps. Instead, they should watch closely and be prepared to make a move —if the system makes economic sense, and if consumers like it.