Scalable Automated Checkout.
Close to 50% of customers will purposely avoid a retailer in the future if they had to wait in line for longer than five minutes (source). That's where Focal's Autocheckout solution comes in. The Focal Product Recognition Camera Camera retrofits your registers to eliminate the scanning step using Deep Learning Computer Vision. This accelerates checkout speeds to get your shoppers out the door swiftly with even less labor and shrink!
1. Reduce checkout and self-checkout time by up to 77%
Focal's Product Recognition achieves 99% accuracy on thousands of products. This powers Focal's Autocheckout which eliminates barcode scanning entirely. Focal's Autocheckout system is the new barcode scanner that scans even non-barcode items like produce.
The average checkout time takes 136 seconds. This includes scanning, paying and bagging. Adding 10 seconds of "setup time" in between, this allows 25 transaction to go through per hour.
For less than the price of an FTE, Focal can eliminate the scanning process with its powerful Deep Learning Computer Vision technology which accounts for almost 50% of the transaction time. And since the cashier is no longer using their hands to scan products, they can bag for the shopper while the customer is paying, reducing the entire transaction time to just 33 seconds on average. This allows 110 transactions to occur per lane instead of just 25.
The average self checkout time actually takes LONGER per item since customers are usually slower at scanning than trained associates. In particular, customers do not know PLU codes so they must pick from a long list of vegetable and produce items that take them 25 seconds on average. This leads to huge shrink issues, mis-scans, and wait times. However, with Focal's Product Recognition, we can eliminate the scanning phase completely, even for non-barcode items. This leads to a reduction in self-checkout speeds from 126 seconds to 33 seconds, increasing the number of transactions a single SCO (Self Checkout) can process from 29 to 110.
2. Eliminate mis-scans for produce, 53% of Sweethearting, which means 62% of shrink through checkout.
Consumer reports have found that:
1. On average, 2-4% of total sales through self-checkouts are not scanned
2. 20% of all shoppers admitted to having stolen at the self-checkout in the past
This adds up to billions in bottom line losses. In such a tight margin business, these thefts can put a store in the red and could result in store closings. With deep learning computer vision, we can flag suspected theft situations and record the entire thing on camera bringing theft to a screeching halt.
Methods of stealing today and how Focal stops it:
- “The banana trick”:
The method: Ringing up a T-bone ($13.99/lb) with a code for a cheap ($0.49/lb) variety of banana.
How Focal Helps: Computer Vision can not be fooled by this. A T-bone and a banana do not look even remotely alike visually! Our algorithms flag this immediately and alert an auditor to come over and stop the theft.
- “The pass around”:
The method: Skipping the scanning and putting the product right into the bagging area.
How Focal Helps: We track hands and when an item enters the bagging area that has not been scanned we send an alert to the user and an auditor to come over and stop the theft.
- "The organic trick":
The method: Bag the organic produce but select the regular tile on the SCO or enter the regular PLU.
How Focal Helps: Organic does in fact look different than regular products as subtle as they may be. Our algorithms can tell the difference between hundreds of fruit types. We can also detect the "organic" sticker on the produce. We automatically input the correct PLU on the customer's behalf to make it impossible to do this.
- “The switcheroo”:
The method: Peel the sticker off something inexpensive and place it over the bar code of something pricey.
How Focal Helps: Our computer vision only cares about what the product looks like visually, not what the barcode reads. We detect this very easily and flag this to an auditor to stop the theft.