PREMIUM CONTENT

Case Studies

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.

SEND IT YOUR WAY

Newsletters

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.

RFID JOURNAL EVENTS

Newsletters

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.

ASK THE EXPERTS

Simply enter a question for our experts.

View Answers to Previous Questions >>

What Retailers Should Test for in a Proof-of-Concept

September 26th, 2016 Retail - Features
By Bob Violino
It's not just about the technology. Companies must also consider the business case, change management, the data and other issues.

As we discussed in our story, How to Conduct a Successful Proof-of-Concept, this important test will help you determine how best to use radio frequency identification to solve your problems. The first step for any company in any industry is to develop a business plan and form a cross-functional team.

Then, you need to conduct a test at the site where the technology will eventually be implemented and used, such as a factory, store or warehouse. That way, you can determine whether a particular RFID solution will work in that environment and address your business issues.

But what, specifically, should retailers consider when conducting a proof-of-concept (PoC)? We asked the experts who have worked with retailers worldwide to develop successful RFID deployments.

Common Shortcomings
Let's begin by discussing what you do not need to test. "The technology works, so retailers do not have to spend time trying to determine if it works," says Bill Hardgrave, the dean of Auburn University's Harbert College of Business and founder of the RFID Lab, a research institute focused on the business case and technical implementation of RFID and other emerging technologies in retail, supply chain and manufacturing. "However," Hardgrave says, "they do need to determine how it works for them and the best portfolio of technology needed to solve their problems."

But often, retailers approach a PoC by first choosing the technology and then determining what use cases it will solve. "This is backwards," Hardgrave says. "The first question asked should be: 'What are the primary problems we are trying to solve?' Then, choose the technology to solve those problems," he advises. "RFID is not plug-and-play, and one size does not fit all."

Another common mistake is not focusing on the major retail use case. "Retailers should always start with inventory accuracy as a major use case," Hardgrave says. "Sometimes, retailers get distracted by uses cases that have the most visibility, such as enhancing the customer experience in the dressing room." While the visible use cases can be important, he explains, the fundamental and foundational issueinventory accuracymust be solved first. The RFID Lab has proved that RFID can increase inventory accuracy from an average of 65 percent to more than 95 percent.

All other benefits of RFID come from inventory accuracy, adds Dean Frew, SML Group's CTO and senior VP of RFID solutions. "When inventory accuracy is measured with a high level of confidence, it frees up the organization to make process and technology change decisions," he states. "Most retailers really don't know what their inventory accuracy is, because it is not practical to measure frequently."

Increased inventory accuracy is the primary use case for RFID in the store, says John Richmond, Tyco Retail Solutions' director of global professional services, since it enables all other benefits. "Simply by knowing precisely what they have and where it is located, a retailer can increase on-floor availability, reduce out-of-stocks, avoid excessive markdowns and lower their inventory carrying costs," he states. "Achieving inventory accuracy also increases the precision of key upstream systems, such as merchandise planning and allocation, and it lays the foundation for secondary use cases like fitting room analytics and omnichannel fulfillment."



So what should retailers test for while conducting a PoC to determine whether RFID can increase inventory accuracy in their stores?

Testing Tags and Readers
Most RFID pilot programs should begin with testing RFID tags with a retailers' own merchandise, says Sharma Harihara, solutions architect at RFID provider Checkpoint Systems. Virtually all retailers using RFID to track clothing, footwear and accessories are using passive ultrahigh-frequency (UHF) technology based on the ISO 18000-6C standard.

GS1, an organization that sets standards for tracking products in retail and other industries, and the RFID Lab's ARC program have developed consistent standards for apparel and footwear tags, Harihara says, so a single tag might be well-suited to multiple merchandise categories, such as denim and polybagged apparel. This makes it simpler to source and test the tags on your merchandise, he adds.

"We are also working with several retailers to tag cosmetics, eyewear and other high-value, complex SKUs," Harihara says. "When we do, we need to plan for additional testing time to ensure RFID-tagged items with metal and/or liquid content are read consistently and accurately."

Tyco recommends that retailers conduct a well-defined pilot, to help them develop and refine the essential processes that can be repeated during a rollout to multiple store locations. This approach also sets proper expectations with both executive sponsors and store teams. "While pilot programs can be run in various ways, we have found those that adhere to a specific sequence of activities will achieve success," Richmond says. "Each is a milestone, and the retailer cannot proceed to the next activity until they've mastered the current one."

First in the sequence is tagging accuracy, Richmond says. Category selection and tagging methodologies are critical, he notes, in order to ensure all items are tagged, the tags are encoded properly, the types of tags are suited to the various product categories and the cost of tagging is minimized. Retailers, he adds, should recognize that tagging strategies will likely evolve as the retailer moves toward full rollout.

Another aspect of testing has to do with the environment in which the RFID tags will be read. Checkpoint conducts extensive testing of RFID-tagged merchandise for a variety of use cases, Harihara says: handheld cycle counting and point-of-sale applications, overhead and zonal inventory systems, and fixed readers in RFID-enabled fitting rooms and back rooms. It also conducts testing of high-volume, high-speed receiving, bulk encoding and verification applications.

"Selecting RFID readers depends on where a retailer's read points will be located," Harihara saysfor instance, at a store entrance, loading dock or fitting room, or on the sales floor. Other considerations, he adds, are the RFID use cases in which the readers will be used. "A store associate locating items for a shopper may use a handheld reader, while a hands-free zonal system may be used in high-traffic areas where inventory needs to be constantly monitored," he states.



In a distribution center, a retailer might use fixed dock door portals or high-volume RFID tunnels for shipping and receiving. "In any application," Harihara says, "accuracy and data quality are paramount, which is why we build device monitoring and data compliance processes into the RFID software."

RFID readers can quickly generate terabytes of data, Harihara says, and are often integrated with systems of record. As such, it is important to test data feeds at various integration points before the system goes live.

Addressing Change Management
Once the tags and readers have been tested in various environments, the next step is to address change management. That's because most retailers rely on store associates equipped with mobile devices for data capture, Tyco's Richmond says. This involves preparing and training store personnel to adapt and become proficient with the new technology and business processes that arrive with RFID.

"Retailers must ensure that cycle counting is being performed on a scheduled basis and in a timely manner," he states. "Training must be available to teach new employees and help improve those with poor performance or inefficiency." Managers require easy-to-use compliance tools to not only ensure good counting, he addsbut more importantly, to understand that associates are using the data to drive operational improvements.

"We work closely with retailers to incorporate testing into change-management processes, to minimize risks at the design phase vs. encountering them at the rollout phase," Checkpoint's Harihara says. "For example, planning how the system will interact with sales associates' daily in-store activities and mobile devices is key to determining the ongoing success of an RFID program. Pilots generally get a lot of attention, but without active program management, many important details can get lost in the midst of daily operations."

Understanding the Data
Many retailers do a poor job of analyzing pre- and post-RFID data, so they don't realize the value RFID can deliver. "Some retailers have installed RFID and then expect miracles to happen," Hardgrave says. "To properly understand the value of RFID, a retailer should look at pre- vs. post-RFID and test vs. control stores. We have seen some retailers make mistakes or poor decisions, because they did not set up the test properly" to enable an accurate analysis.

Analyzing the data is the third step, Richmond says. To validate an increase in inventory accuracy, he explains, retailers must compare RFID inventory data with data from manual bar-code counts and information from the retailer's enterprise inventory system of record. A big mistake retailers make is a failure to proactively and carefully identify control stores so data capture can occur in parallel with RFID pilot stores, he says.

Next, Richmond says, retailers can "measure the success of the RFID pilot program," to determine the benefits enabled by the new levels of inventory accuracy. "A carefully selected control store is matched with each pilot store so that results can be captured in parallel," he states. "Companies should consider standard retail metrics such as sales uplift and gross margin, and omnichannel metrics such as conversion, pick performance and cost."



To measure the benefits of improved inventory accuracy, Checkpoint's Harihara says, retailers first must establish key performance indicators (KPIs), such as shelf availability, same-store sales or omnichannel fulfillment accuracy. The KPIs are defined by a steering committee that represents multiple functionstypically, IT, operations, inventory control, supply chain and merchandisingbefore a PoC can be conducted.

It's important to note that even after inventory accuracy is achieved, there can be a delay until retail business metrics start to move, Richmond says. "The retailer's upstream systems need time to react to the new accuracy," he explains, "and for the new merchandise allocation to work its way through the supply chain down to the store to drive down out-of-stocks, for example."

Inventory accuracy can be measured without integrating data into back-end systems, but during a PoC, retailers might want to test how an RFID system integrates with back-office systems, such as enterprise resource planning or warehouse-management platforms. "Oftentimes, there is a thought that integration will be too hard," SML's Frew says. "It actually can be quite simple and be leveraged into significant value during a pilot, and is essential to realizing value." Integration should be assumed in the business case prior to conducting the PoC, he adds.

Automating integrations so RFID data can be utilized within legacy systems is often a critical component of pilot programs, Richmond adds.

From PoC to Roll Out
Retailers should begin with a PoC and the associated testing in one or two stores, Hardgrave says, to determine how best to deploy the technology and uncover unanticipated issues.

"The speed of retail keeps increasing, and there is often a huge push to roll out an RFID program before a peak selling season," Checkpoint's Harihara says. "Deployments can happen quickly, but it's important not to cut corners on the planning or testing process.

"Successful RFID programs are holistic," Harihara says. "They encompass the individual component parts of the RFID system and how retail employees will use the system in everyday operations. Because all RFID components are interconnected, a single change to a reader, use case, tag format or merchandise category can change how the system performs. We are firm believers in testing thoroughly beforehand and providing ongoing monitoring of the system once it is installed, from device performance to cycle count accuracy, event logs and process compliance."

A common error retailers make is to "test once and run forever," Harihara says. "The retail environment is always changingdifferent merchandise, new store formats, new sensor devices, new firmware upgrades, new use cases and applications." All can impact the quality of RFID data, he adds, so it's important to incorporate periodic data, usability and system audits into each implementation plan.