There are many different ways that RFID can be deployed in a retail environment, including these four:
- Handheld Readers — Retailers tag certain categories of items with RFID labels and then the store associates walk around and take inventory using handheld readers.
- Exit Monitoring — Fixed readers are installed at all entrances/exits, including the dock doors and the doors between the back store and the sales floor. The readers detect when items enter or leave the store, and when they transition between the back and the front of the store.
- Whole Store/Area Illumination — Readers are placed throughout the whole store or an area of the store, often mounted in the ceiling. These continuously read the items in the store. They are often RTLS (real-time locating system) readers, which can not only tell you that an item is in the store, but where in the store each item is located.1
- Smart Shelves — RFID readers are integrated into the shelving or fixtures holding the merchandise. This has the potential for extremely precise locating and counting, but at a high cost.
The vast majority of retailers implementing RFID have chosen to use handheld readers, primarily because it has the lowest required upfront investment. Does this mean that all retailers should be using the handheld reader approach, or do the other approaches have a natural home within the universe of retailers? We have seen retailers piloting, and in a few cases rolling out these other approaches. Let’s take a look at how these four different approaches compare.
Handheld Reader — Economical with Compelling ROI
Inventory counting using handheld RFID readers is much faster (25X or more) as well as more accurate than manual inventory counting. Because counting takes much less time, retailers can count the inventory much more often, typically weekly or every other week, rather than the usual practice of once or twice a year. Combined with the inherent accuracy advantage of RFID, this results in dramatic improvements to inventory accuracy, often from the neighborhood of 80% accuracy (pre-RFID manual cycle counting) to 98%-99%+ using RFID. The subsequent drop in out-of-stocks and rise in sales is the key driver for RFID implementations, as discussed in “Understanding the RFID Renaissance.”
However, handheld readers have some drawbacks. For one, you are still relying on people to go around and scan the inventory items. People make mistakes. Or they forget or miss areas. Most of the retailers with whom we spoke mentioned how critical it was to have good change management, training, and compliance monitoring, to make sure that store associates do a good job. Another weakness of the handheld approach is that it does not provide continuous monitoring. This allows inventory accuracies to drift a bit between counts and also does not provide immediate notification when there is a theft.
Exit Monitoring – Adding Loss Prevention
Exit monitoring can eliminate the need for hand scanning of tags (though some retailers may choose to do both). In contrast to the weekly or bi-weekly updates that handheld readers provide, an exit monitoring system continuously monitors all entrances and exits, providing a nearly real-time tally of the inventory as items move into and out of the store. This provides the additional functionality of aiding in loss detection and loss prevention. If you compare the inventory going out of the store with that being rung up at the POS system, you can detect when items have left the store that were not paid for. Alternately, tags can be written when paid for, so you directly read whether an item is leaving without being paid for. To reliably sense whether inventory is entering the store vs. leaving the store, these systems generally need some type of direction detection. This can be achieved by using auxiliary technologies (such as infrared or video) or by using RFID readers that have locating capabilities (for more on this, see section below “Whole Store/Area Illumination”).
Exit Monitoring vs. Traditional EAS
Using RFID exit monitoring for LP (Loss Prevention) has a distinct advantage over traditional EAS systems: you can detect exactly which items were stolen.2 That way the store staff knows to replenish those items. Further, they know which types of items are being stolen and can focus loss prevention strategies and deterrence efforts on those items. RFID may not perform as well as traditional EAS systems in some circumstances, such as ‘body shielding.’ For that reason, some retailers chose to keep their existing EAS system and augment it with RFID, in which case they may justify the RFID primarily based on the sales uplift, rather than just LP.
Because of the emphasis on loss prevention, these exit-monitoring RFID systems are often provided and installed by one of the established EAS vendors, Tyco-Sensormatic or Checkpoint. These vendors have decades of experience in exit monitoring and have some tools that can help in thwarting would-be shoplifters attempting to overcome RFID’s weaknesses, such as devices to detect when a customer tries to bring a foil bag into the store.
Inventory Accuracy Improvements
Like handheld RFID, exit monitoring also provides an increase in inventory accuracy over traditional manual cycle counting. Exit monitoring may provide slightly better inventory accuracy than the handheld reader approach. It depends in part on how well the handheld reading is being done. It will almost certainly provide an improvement over poorly trained store associates whose inventory taking performance is not being closely monitored. If employees are doing a good job, the improvement is likely to be smaller, particularly if the accuracy using handhelds is over 99%.3 In any case, exit monitoring eliminates the need to rely on a person to do the reading and to do it properly.
There are physics challenges which can prevent exit monitoring systems from achieving 100% read rates. For example, if a customer has a package tucked under their arm while they are leaving the store, there is a good chance some items might not be read.4 Shoplifters may put items in a foil-lined bag, which makes those items nearly impossible to detect.
Compared with a handheld solution, exit monitoring has a higher initial cost of the infrastructure (fixed readers cost more to buy and install, compared with handhelds). However, handhelds incur ongoing labor cost associated with manual scanning of the inventory. The fixed readers do not require that manual labor.
Whole Store/Area Illumination — Continuous Real-time Visibility + Location
Multiple fixed readers can be positioned to provide complete coverage of an entire area or the whole store. Usually this setup incorporates locating technology such as that found in Impinj’s xArray, Mojix’ STAR 3000, and RF Control’s ITCS.5 When the whole store is illuminated6 in this way, it provides a continuous picture of what is in the store and where it is in the store or back store. This location information can provide some additional benefits. The system can detect when items have been misplaced, such as when a customer moves an item. It can help the store associate find items in the back store.
These systems do have their limitations. Products that are stored tightly together, contain metal or liquids, or have metal/foil in their packaging are likely to have issues. This weakness is the case not just with whole store illumination, but with all of the methods described here. Whole store readers have the added disadvantage of greater distance from the products. However, because these readers are continuously monitoring everything within the store, they have more chances to read tags as inventory is moved through the store and placed onto the shelf, and so may be able to use ‘last known location’ to infer where inventory is, even when those items are unreadable as they are tightly packed on the shelf.
In addition to the added location benefits, whole store illumination can provide additional loss prevention benefits. For example, if a shoplifter places an item into a foil-lined bag, the system can alert store security personnel that an item just ‘disappeared.’7
Few retail stores have used this technique of whole store illumination, largely because the cost of the fixed reader infrastructure is much higher than either the exit monitoring or handheld reader approaches. Those stores that have implemented whole store illumination tend to be small-format, higher-end apparel.
Smart Shelves — Ultimate Precision, Works with Difficult Products, but at a Cost
Smart shelves incorporate antennas into fixtures, drawers, and display cases. By getting very close to the products and surrounding them, they are able to provide higher performance, including working with more difficult materials and providing more precise locating capabilities. To date, there have been very few retailers rolling out smart shelves across their stores.8 We have some interest from jewelers who want to keep very close tabs on inventory moving in and out of the display case. But overall, smart shelves are the most expensive of the four approaches we outline in this article and have seen little uptake beyond pilots.
Which Approach Is Right for Different Retailers?
The fact that most retailers have chosen to use handheld readers does not necessarily mean the other approaches have no future. Handhelds have been a great choice for pilots, trials, and rollouts of specific categories. They don’t require much capital investment and provide a lot of flexibility, since they can easily be moved from one part of the store or another, or to a different store. Fixed readers are more expensive to install and are by nature relatively fixed in position (they can be moved, but not without investing additional time and expense).
Some of the retailers with the most RFID experience have decided to tag everything in the store. This includes not just apparel-only retailers, but even some general merchandisers, such as Marks & Spencer, which announced in 2012 that they intend to tag all general merchandise in addition to all apparel. Once you are tagging everything in the store, it might have implications for which method makes the most sense. In addition, there are some categories of products that are hard to read, even with a handheld reader, such as densely packed cosmetics with lots of metal and liquid content. Perhaps for those categories, a smart shelf approach might make sense.
Cost-Benefit of Switching from Handheld to Exit-Monitoring or Whole-Store
There is a potential benefit in greater inventory accuracy when switching from handheld RFID to exit-monitoring or whole store illumination. Exactly how big that benefit is has yet to be proven out across retailers. The inventory accuracy we are hearing from most retailers using handhelds is in the range of 98%-99%. That does not leave much room for further improvement. The value of increasing accuracy by a percent may not justify the additional investment. However, when combined with other benefits, such as loss prevention and/or possibly insights into customer shopping patterns, it could make a difference. The size of the store, density and value of the merchandise, and the mix complexity (size, color, style) will also impact the investment vs. benefit per square foot. A store with a small footprint, high-value items, and high mix complexity might justify whole-store illumination.
These next couple of years will be interesting to watch the evolution of approaches used for retail RFID. We expect to see a diversification of product categories being tagged, beyond apparel, as well as more stores tagging 100% of merchandise. Eventually some retailers will take the plunge into doing RFID checkout as well. All these factors could change the balance to a more diverse mix of RFID store reader strategies and approaches, different from today’s heavy reliance on handheld readers.
1 Systems providing RTLS functionality for passive UHF systems typically use an array of antennas, measuring the phase difference in the signals seen by each antenna in order to provide location. The location precision varies depending on the system and environment. — Return to article text above
2 Traditional EAS will detect that an item was stolen, but not which one. — Return to article text above
3 At least one chain is achieving inventory accuracy of nearly 99.5% using handheld readers. — Return to article text above
4 UHF radio waves do not penetrate water well. Since the human body is largely water, it tends to absorb and block the signal. — Return to article text above
5 These systems have some similarities and some differences in architecture and performance. The solution providers make a range of claims about distance and location accuracy. — Return to article text above
6 The term ‘illumination’ in this context refers to illuminating the store with radio frequency radiation, not visible light that can be seen with the human eye. — Return to article text above
7 One caveat: Other circumstances can cause a tag to suddenly ‘disappear’ (i.e. not be readable), such as when items are too tightly stacked in a shopping cart or some metal products or liquids are put near them. We have not seen any studies of how well this technique works in practice, i.e., whether too many false alarms would be generated in real-world use. — Return to article text above
8 A few years ago, there was a fair amount of interest in smart shelves. One of the early pioneers was Vue Technology. Vue was subsequently acquired by Tyco, who has primarily leveraged Vue’s software platform, rather than their smart shelf technology. — Return to article text above
To view other articles from this issue of the brief, click here.