Apr 18, 2004y name is Christine Overby, and I am a senior analyst at Forrester Research, a market research company that looks at the impact of technology on business, I have spent the past three years looking at this really exciting our RFID market. Before I get into the compliance study I wanted to share a story with you. I was at dinner last night with a colleague of mine at a nearby restaurant, and we happened to have the program for this event on the table—just kind of casually laying on the table—and a waiter walked by and he saw it, and he started talking to us about using RFID tags to track cattle for mad cow disease. He thought that this was really interesting.
That’s just one data point, but it’s the data point of the mainstream public that now knows about RFID. In fact, just half an hour ago, I was walking through the conference center, and I saw this man and his son dressed as tourists and the father was telling his son about RFID. It’s really just amazing. I think what it underscores is that we are here not to talk about the vision. That vision is starting to be understood certainly by this group but also by the public at large. (Download presentation from Forrester's site.)
So what I want to talk about is how we get to that vision, the cost. And that’s why I am going to spend some time on this research that we did—RFID at what cost. Mark had mentioned we did a three-month study. We talked to dozens of suppliers and dozens of vendors. A lot of you are in the audience, and what we learned from that—what I would like to share with you —is that Wal-Mart compliance is about learning the value of RFID and controlling the cost.
I hope to share with you suppliers in the audience what you need to do between now and January 1 of next year. So we will do that by taking a look at three questions:
What’s the reality of Wal-Mart’s January 1, 2005, mandate?
What is the cost and the ROI of compliance?
Then finally, how should you respond?
If we take a look at that first question, we really have to look back at Wal-Mart’s strategy because RFID is not an aberration for Wal-Mart. This is a part of a timeline of Wal-Mart’s technology investment from 1962 all the way to present. What you will notice is that in the 1960s, they had the first computer to support operations, and they had an inventory management system. They were years ahead of the competition here. In the 1970s, they invested an electronic point of sale and satellite networking to connect their stores. In the 1980s, it was about the UPC, the bar code and EDI, and then they built on that in the 1990s with Retail-Link and CPFR.
During the dot com days, they did focus a bit on shopping experience technologies, like Wal-Mart.com and self checkout, but they returned to their roots in 2000 with the [Auto-ID Center] town test the UCCnet mandate and then ultimately the RFID mandate. Now the reason that I bring these up is that these technology investors really focus on the supply chain. Mark mentioned this morning that Wal-Mart has a brilliant and focused strategy that centers on their everyday low price model, and RFID in this sense is about delivering on that strategy and it works for them.
This is some research that Forrester did last year with 4,000 US consumers. We gave them about a dozen criteria for shopping and asked them what they appreciate most about Wal-Mart. And these are some amazing statistics. 86 percent say that Wal-Mart has great prices and product assortment and 76 percent say that they have a good amount of in-stocks product availability on the shelf. These benefits they really center on supply chain excellence and that’s what I want to reinforce—RFID is about building on that supply chain excellence.
So lets take a look at the mandate itself. I am not going to spend a time here because I know that many of you in the audience are getting details directly from Wal-Mart. But we look for the details too at Forrester and there were three that really stood out to me. They want deployments by January 1, 2005, in three Texas distribution centers. I think this is important because it's deployment and not pilot. They want 100 percent readability on cases and palette. And then finally they want strategic thinking.
I was at the NRF conference earlier this year when I stood in a room about this size and [Wal-Mart] said to the audience that slap and ship is quite disappointing. Wal-Mart said that this has to be about a return. It can’t be a cost of doing business. As part of the study, we took Wal-Mart at its word and we really first tried to understand where the benefits were, the promised benefits of RFID. We started with the benefits that the original Auto-ID Center put out. And I am not going to go through all of these but I’ll take the top four.
The Auto-ID Center had said that inventory management and out-of-stocks are among the biggest returns, and we think they are right here, by the way. Its really unquestionable: RFID is going to provide an amazing amount of visibility in the supply chain. But I want to talk to you about what it takes to get to inventory management and out-of-stock because to really achieve all of these benefits to get, you know, all of this bar [on the presentation slide] you have to implement RFID the across the entire supply chain.
This is strategic supply chain ROI. When we talk to a lot of manufactures, that means source tagging. It means tagging at the manufacturing facility, cases and palette. Forrester does not think that this is going to be something that happens in the next twelve months. We have talked to people who have said that this is just simply too much process reengineering for the near term. I was actually speaking with one manufacturer who said that in order to do source tagging, they would have to implement RFID at over a hundred finished goods facilities, and they would have to segregate the inventory across their supply chain. So while we do believe that this is going to be a benefit of RFID, it will be a benefit that will occur in the two to three year timeframe.
So lets a take look at theft. We talked to people who said you could start to achieve this if you implement RFID in a single distribution center. But again here to get to that full bar you need to implement theft or you would need to implement RFID infrastructure across the supply chain. Now many of you guys know that part of the supply chain is outsourced. In fact 80 percent of personal care manufacturers used co-packers, so what this means is that all those guys need to implement RFID infrastructure as well. So we talk to those folks, and they have told us a pretty similar story. They are in the same boat that many you suppliers are. They are thinking about how they implement RFID and how to roll it out more broadly, so this is another benefit that we believe is going to be beyond the scope of the January 1 deadline.
Let me talk about the last one—warehouse management. This is the area where Forrester sees the most immediate return, and this is because there is a lot more accuracy and a lot more granularity that you can get when you match your advance shipping notification with a Wal-Mart receipt. Here it really becomes a question about the break-even point. We think that the break-even point for warehouse management ROI is also beyond January 1, 2005, and the problem here is not really an RFID problem, it’s a data synchronization problem, because for this type of data to go back-and-forth you need standards between the supplier and the retailer. We just did a study of safety manufacturers and retailers, and we found that 70 percent are still struggling with internal product data.
These are the four benefits, and the reason that I really call them out is I want to point out that while they are benefits they are benefits that are going to take multiple years to achieve. Now many of you in the audience have been asked to use these benefits to build your business case. So it kind of brings up the question can Wal-Mart get what it wants by January 1, 2005 and Forrester thinks that the answer here is, no. Many of you have seen the New York Times article I heard actually that a couple of you were passing it around yesterday and so there is already talk that the deadlines are moving back and they are being redefined. Here I think—you know I am not trying to bash Wal-Mart—but I think that its important to actually recognize them for drawing the line in the sand because this deadline has gotten the industry to move and I think its unquestionable that Wal-Mart has taken a risk here in putting themselves out and really articulating this. But we believe that right now there are 277 days until January 1, so we need to focus on what happens now.
So let me get on to that. Some of you in the audience might be looking at my promised ROI model and saying we really don’t need to do something. I want to be clear here because we believe that you still need to do something because Wal-Mart is a very important customer for many of you and RFID will be a very important technology in your supply chain. So let me take you through what we think that you should do and if there are three things that you remember from my talk today, these would be them.
The first is to start with the slap and ship approach. This might be a little controversial. Let me define what I mean by slap and ship. This is applying tags and really focusing on implementing reader infrastructure and other infrastructure and the distribution centers that service Wal-Mart's Texas DCs. Now, it's really critical because even with this straightforward approach, you will be doing a lot of work. But the important thing here is that with this approach you begin to focus on learning how RFID works.
I have talked to Schweppes, Procter & Gamble, Scottish & Newcastle many of the early adopters of this technology, and I think Mark even alluded to this, this morning. Many of them have said that they really didn’t know where the benefit was going to be. They had an idea, but they didn’t know where it was going to be until they actually implemented the technologies. So when you start to test RFID and when you start to really work with it, you’ll begin to understand how it will deliver the eventual ROI to your business.
At the same time you need to minimize the amount of distribution center reengineering. You will have some reengineering, and I will walk through what a typical scenario would look like, but here you don’t want to rethink your entire warehouse operation processes because you will be focused in the near term on learning how the technology works and really starting to understand how it impacts the business. So let me talk about how compliance works.
When I was doing this research and I kept speaking to people, they said: "You know, Christine, you really can’t talk about cost or talk about benefits in the abstract here. You’ve got to define a company and work backwards." So we did. This is a company we created, called XYZ Manufacturing. I am going walk you guys through the profile of this company so then we can talk about the cost that they would occur. They have $12 billion in North American sales, and they are food manufacturer. The reason that I picked this number is because it represents about 25 percent of the top 100. So it's pretty typical. They have three distribution centers that ship to Wal-Mart 's three Texas DCs and this was out of the total distribution network of 20 DCs.
Obviously, this is assuming a geographic centric distribution approach versus the product centric approach, but we picked one that we have seen companies adopt. They ship 15.6 million cases annually. This is based on an average case price of $20.00 and about 13 percent of those shipments go from the manufacturer’s distribution centers to Wal-Mart distribution centers. Again it’s fairly typical of the food industry.
Okay, so here is what we believe that XYZ would do. They would first start by manually tagging the cases during picking and packing. This allows them to figure out how to orient the tag so that they can improve readability. Then after that, they verify the case tags on the sortation conveyors, and this is using fixed conveyor readers. In this scenario, what they are picking up is the defective tag, and the reason that I chose that to use fix conveyor readers versus handheld is because we hadn’t seen much more than UHF prototype available in the market, and I am actually really interested to hear some of the DOD guys talk about how they had to go to some of the handheld vendors and have these technologies made.
Then they are going to segregate the inventory in the shipping zone. This allows them to identify those Wal-Mart orders, make sure those orders aren’t leaking into other customer’s orders. But again, they are not rethinking the entire distribution center at this point. Finally, they are going to read the palette on the way out.
Here is the dock door reader that is verifying the palette EPC and associating it with the advanced shipping notification. You guys know that this is a major stipulation for Mal-Mart, and so this is how XYZ chose to do it. Okay, so that’s the broad approach now let me take you through the specific costs.
Okay, $9.1 million in year one. This is the upfront investment and operating cost for 12 months. Now that’s a lot of money. I’ve tested these figures on many of you in the audience, many of the vendors to make sure that we weren’t overstating this. But let me take you through the specific line items to talk about each briefly.
Tags: $7.6 million. This is the biggest area of investment, and here we assumed a tag price of 40 cents. Hardware was $329,000.00. This is for both servers and readers. We assumed a fully loaded cost here between $8,000 and $10,000. This is inclusive of the software, the box, the antennas, the cabling, the power and the installation. We have actually heard some interesting stories about how the installation can be a big hidden cost. Companies have had to move fire hydrants and shrink-wrappers that were in between the dock doors in order to accommodate the readers and so that’s one of the reasons that it costs more than the simple reader.
For software we came up with the figure $183,000. This is assuming basic middleware to manage the readers and a database to contain the EPC codes. This is not a broad application integration into your backend system. For consulting and integration, the cost was a $128,000. This is based on an hourly rate of $250 for an RF Engineer, which sounds high but what we found is that because of the demand in the market and because there are so few people who know how to implement this right now, that's what many of the consultants were charging.
An internal project team cost of $315,000. Here it’s a mix of full-time and part-time resources. This includes full-time people from IT and logistics and part-time people from customer service, packaging and finance. This is based on a fully loaded annual cost for these employees.
The tag and the reader testing is $80,000. Here’s a line item that a lot of people forget about. But what we heard when we talked to the early adopters is that they need to test all of the product types they have with the tags with the reader technology that they are using. Here is another big one: Additional warehouse labor, and this is almost $500,000. This was based on the cost of applying or the time that it took to apply these tags for slap and ship. This means around two to three people per shift at this distribution center and then we have $39,000.00 for training for both new and existing employees.
You will notice that some of these costs bear a disproportionate weight on the total cost. It really underscores what the industry still needs to do to get to the RFID vision. The tag makes up 80 percent of the cost and prices aren’t dropping quickly. We hear a lot of talk about volume and cost but it actually takes more than just volume to get these tag cost down. It takes standardized form factors. It takes aggregated orders and it will take tag production innovation.
Professional services run high and the cost will only increase. Again what we expect is that the collective mandates have a kind of unintentional affect they keep the IBM and the Accentures of the world very busy and will keep these costs high in the near term and then finally the case tagging right now adds rather than decreases labor. This is a probably the most counterintuitive finding in this research and what it really does speak to is that if we are going to get to the vision of RFID as a labor saving technology we have to figure out how to integrate this into production and integrate this into the process.
So, how can the vendors help? For you vendors in the audience, you have a responsibility to address those three challenges first. I know that you’re probably looking at the cost model and there might be some glimmers in people’s eyes, but this is an area where success is going to breed success and failure is going to breed failure. If you don’t, as a vendor community, address these problems then there is a danger that these won’t get off the ground.
So here is what we think that you can do. Create the guidelines for case tagging. I was speaking with a guy who was running RFID for one of the pharmaceutical companies and he said: "The problem right now is that I’ve got all of these tag vendors coming to me and asking me what tag I want and that’s backward. I don’t know what tag I want. I need them to tell me the tag that I need."
What this underscored to me was that the users, the suppliers, still need basic help to understand things like tag performance and where to place the tag on the case. Also you should focus on source tagging infrastructure for us to really get the benefits of RFID. This needs to be across the supply chain and that means RFID tags that are in line with the manufacturing process. There are a couple of different ways that this can happen.
In the near term, the integrated printers and automated applicators and eventually the RFID tags that are embedded in case-packing material and then finally to improve the reader middleware application interface. This is the area where users tell us that the market is most confusing. There is overlap, and it’s not that you vendors are trying to pull the wool over the suppliers’ eyes, but everybody’s involved right now and it’s really an emerging market.
It’s really about focused partnerships, dedicated partnerships, shared research and development. A lot of people use the term "hardware agnostic" or "technology agnostic" but we think that it’s really the responsibility of the application vendors and the middleware vendors to go and understand and build the adaptors to the specific hardware. So it’s not actually about being hardware agnostic. These partnerships are really critical because here’s what the market looks like today. We have got tags, hardware, data management applications and we have got a patchwork of new and existing companies that are going after the market.
Phillips and Texas Instruments are among others going after tags and labels. Intermec is going after hardware and readers. Some of the new companies like Matrics and Alien are going after both. Cisco is getting into the data management and the networking part of it. Sun, IBM, OAT Systems and ConnecTerra are going after data management. The big enterprise app vendors who have the middleware play, like Oracle and SAP, are going after apps and data management together and then finally the specialized vendors like Red Prairie and Manhattan Associates are going after the application layer with a little data management as well.
Now this is the reality. This is an emerging market. There are a lot of players right now, a lot of heterogeneous players that are offering solutions in this market. So here’s what we think that you suppliers in the room should do given the reality of the market. I worked with the easiest product categories first. Do not wait for the market to resolve the problems of liquid and metals. Smart people are working on this. But start with the products that have a minimal amount of these so that you can begin to get your feet wet.
Pool the tag orders with internal and external buyers. This is aggregating orders, both within your business unit but also with suppliers who might have similar product type this will bring down the cost of the tags somewhat. Tie RFID to data synchronization. Right now we have companies like IBM and Trigo that are stating to go after this problem. Work with those vendors to get them to articulate a road map between what’s going on with data synchronization right now and how those data synchronization technologies will accommodate new RFID data.
Set up one throat to choke. I don’t care what vendor you choose for this. It could be tag vendor, it could be middleware, it could be applications but it needs to be somebody who has experience and resources that they are putting into your project and someone who has a willingness to share the risk. This is going to help with some of the readability and the performance problems today.
And finally, create a direct report for RFID to the CEO. One of the other things that we have noticed is that these collective mandates have started to get the attention of the board and shareholders, and so the time is now to have an RFID director for that is talking to the CEO about the cost and how this is going to change the business. I want you to remember three things today. Start with a ‘slap-and-ship’ approach. Focus on learning how RFID works, and then finally minimize the amount of distribution center reengineering this is all about getting us to January 1. Ultimately there is a lot of process reengineering to realize the benefits. Thank you.