In just the past nine months, Amazon soft-launched its first Amazon Go cashier-less retail store, opened three more locations and announced a commitment to 3,000 stores across the U.S.. It is now clear that the convenience and novelty of cashierless shopping is no longer a “Future of Retail” concept, but a consumer shopping “must-have.”
Since Japanese retailers are often ahead of the U.S. in customer service and tech innovation, we suspect that adoption of this technology will be faster and more comprehensive in Japan.
We’re pleased to share market insights from Krishna Motukuri, co-founder and CEO of Zippin, a Silicon Valley-based CVG portfolio company, that is enabling any retailer to offer an economically-viable, cashierless shopping experience.
Q: When and why did you found Zippin? Was your vision inspired by your time at Amazon?
While my background is retail technology, including a stint at Amazon, the real genesis of Zippin came from a shopping experience four years ago. My wife and I were new parents and she asked me to stop for organic milk on the way home from work. When I arrived at our local grocery store the lines were very long and I decided there was no way I would wait for 10 minutes just for milk. I ended up picking up non-organic milk at a corner store instead.
The whole experience got me thinking that we need a better solution. And over the last two years, my now co-founder at Zippin and I devised methods to apply machine learning, AI and visual cognition technology, toward making shopping quicker and easier.
Q: We see lots of start-up solutions working on retail efficiency right now. How might you categorize the various solutions and their advantages?
There is a lot of excitement right now for this kind of frictionless shopping experience. It is terrific that demand from Amazon competitors is fueling innovation and competition amongst start-ups.
Solutions tend to fall into two groups: those relying on single source data systems – typically from overhead cameras, and those leveraging multiple independent data sources, such as cameras and shelf sensors. We believe that two independent data sources leads to the highest level of accuracy within densely crowded retail spaces (as deployed by AmazonGo and Zippin).
Single data-source solution providers tend to argue that they are more affordable for retailers in terms of an upfront capital expense but, in reality, the loss of accuracy will result in lower margins – which are already razor-thin in retail, and can severely impact the viability of such approaches.
In addition, there’s a whole category of somewhat older solutions that rely on smart shopping carts equipped with cameras to track the products in the carts. As far as we know, these solutions have not yet been successfully deployed. Due to the mobile nature of shopping carts, and the numerous ways which customers can put products into shopping carts, the computer vision technology needed to accurately recognize product movement can be more challenging than identifying products picked off static shelves.
Q: What do you think is the biggest misconception held about autonomous shopping right now?
The biggest misconception is that “cashierless” or “checkout-free” retail means “human-free” shopping. What most people don’t realize is that shopping has always been more about discovery than payment. But unfortunately, most stores today have their employees staffing checkout stations – scanning products and collecting payments – instead of helping customers find products they want, or discovering products they might like. Now stores can free up their checkout staff to provide superior customer service.
Q: Which specific retailer categories do you think will adopt the quickest? Who will see the most value long-term?
The earliest adopters of this technology fall into two categories. First, those operating smaller urban convenience store markets with many products and high sales volume per square foot, including airport stores and hotel and office building kiosks.
Second, operators of large retail grocery and drug stores who may be interested in opening smaller “stores within a store” – i.e., a small, dedicated area of real estate within a larger store, stocked with popular items that can be sold 24/7 and purchased in a separate area of the store via autonomous shopping apps like Zippin.
In terms of long-term value, there’s a strong proposition for all types of retailers, from large format stores to the smallest and all in between, especially as the cost of the hardware decreases and the insights from our data systems become clearer.
Q: If you were a Japanese executive working in the traditional retail industry, what advice would you lend?
We are about to experience one of the biggest shifts in retail since the introduction of bar codes. My advice to the CEO of every traditional retailer in Japan is to develop a plan for checkout-free stores. While the benefits to shoppers may seem obvious, what may be less obvious is the superior economics and additional benefits to the retailers themselves:
1) Higher sales per square meter. By eliminating checkout counters, stores can save 10 – 20% of floor space. This space can be utilized to increase product selection and drive additional sales. On top of that, there will be an increase in store visits when customers who were previously deterred by long checkout lines walk in when they don’t have to worry about checkout.
2) Real-time inventory tracking. Checkout-free technology solutions like Zippin work by keeping track of products on every shelf in the store, all the time. As a result, retailers can get real time, up-to-the-minute accurate inventory data for each shelf. If a shelf is empty or running low, store staff (or suppliers) can be alerted. If a customer puts back a product on the wrong shelf, the system can automatically track that and help the store staff find the misplaced product quickly and return it to the correct shelf. These capabilities mean better experiences for the shopper and streamlined operations for the retailer.
3) Advanced analytics. Retailers get detailed insights at the level of individual shoppers, on par with e-commerce, and never previously possible in physical stores.
4) Loss prevention. In a checkout-free store, every customer checks in when entering the store. Once inside the store, any product a customer picks is automatically added to that customer’s virtual cart. When two types of sensors are used (eg: cameras and shelf sensors), the accuracy increases substantially, making it virtually impossible to steal.
Read on Nikkei Site