SCDigest: RFID in CPG to Retail – What Really Happened?
Through
a special arrangement, presented here for discussion is a summary of a current
article from Supply Chain Digest.
Has there been anything much stranger
than the tale of RFID in the consumer packaged goods to
retail supply chain?
The journey has gone from visions of supply chain transformation
to now, literally, the concept being at a near dead standstill — though there
is clearly a connection to today’s very lively efforts with RFID for item-level
apparel.
Walmart is doing nothing with EPC tagging in the CPG area, now focused
on apparel programs. A P&G spokesperson told me recently that as far as
he knows the company has no active RFID projects or pilots currently underway.
This from a company that was leading the charge not that many years ago and
had at one point, I believe, at least 20 people working on RFID. P&G, as
just one example, developed what became the industry standard requirements
document for RFID-capable fork trucks. Even Tesco and Germany’s Metro chain,
which continued ahead after Walmart had clearly started to bail, have done
nothing new for about three years.
Was this vision in the end just full of mush?
Or is the value still there waiting to be reignited?
“There was sort of a collective euphoria back then that probably didn’t
look hard enough at the real business case for consumer packaged goods,” said
Simon Ellis, an industry analyst at IDC Manufacturing Insights and for several
years a “supply chain futurist” at Unilever North America. He noted
that even at five to seven cents per tag, a level we are just getting to now,
that would equal the cost of the secondary packaging itself, “which would
be very difficult to justify” in low margin CPG products.
As the reality
started to hit relative to the cost impact of the Walmart program versus the
benefits, CPG companies started to push back despite such strong earlier enthusiasm.
In
the end, here is what I believe happened:
- CPG companies and Walmart did not do enough “hard math” to really
understand the returns before making substantial investments in RFID. Other
retailers sat back and let Walmart do the “dirty work.”
- There was a “herd” mentality that contributed strongly to this,
supported by the self-interested RFID hype machine in some parts of the media
and consulting community.
- Many CPG companies did and continue to believe there is great ROI in some
product categories, the most prominent being in promotional execution but
also others. But Walmart basically gave up on those (not clear why) and no
other retailers have stepped up. Manufacturers can’t do it without retailers.
- Walmart did not design or execute its program well, especially in having
a mass mandate across hundreds of suppliers, when P&G and others were
arguing for a segmented approach based on value. I believe the value for
these “advantaged” products (P&G’s term) is still there. This
is more the approach Walmart is smartly taking now with apparel.
- Venture capitalists lost hundreds of millions funding RFID companies when
the Walmart bonanza never happened, but those investments moved the technology
rapidly forward, enabling today’s apparel programs and more.
Discussion Questions
Discussion Questions:Â What do you think happened in RFID for the supply chain from consumer packaged goods to retail? Is the value still there?
RFID technology hit the treadmill after the economy went south and other priorities fell into place. In addition, RFID ran into a wall when retailers became concerned about consumer perceptions with regards to how it will be used.
I totally agree with Simon and Dan. Working with CPG companies during the Walmart mandate, everyone kept saying where’s the ROI? The only reason many started investigating RFID was because of the Walmart mandate. RFID has an application for item level tracking for expensive items, electronics, pharma, asset management like beer kegs, satellites, etc. Unfortunately, for a box of soap it doesn’t add much value. A barcode gives you all the info you need on that item.
As for some trials that wanted to go to an automatic check out where everything was tagged in the basket and it went through a reader tunnel, and voila, all was checked out, it was a good concept. But as Simon said, the cost of the tag and less than 100% accuracy of the technology are hold backs.
We have clearly moved past the hype phase and “trough of disillusionment” (Gartner term). With a little help from a tough economy, the industry has refocused the vision and activity around incremental value areas. The most notable example is the apparel activity. Look for renewed investment and momentum building on success in these areas and also related technology evolution (i.e interactive NFC).
I remember even during the height of RFID mania in the mid-2000s there were a lot of significant technical problems, such as RFID-tagging containers of liquid and dealing with metal packaging, that as far as I know were never resolved. It’s too soon to count RFID out, but I think it’s safe to say we are at least a decade away from seeing the kind of 100% transparent RFID-enabled supply chain everyone was dreaming of in 2005.
Mostly I think Dan has it right, but I think the sequence of events was a bit different. I saw this article last week, so I challenged myself to explain the situation in 100 words or less. I ended up with 127:
1) After the dot com bust, VCs were really looking for the next big thing to put their money into
2) Along came RFID and a lot of UK/MIT academic enthusiasm (and many analysts were complicit … because THEY needed something to talk about after the bust–it was the only game in town).
3) Walmart’s CIO bought into the notion that they could remove one body from every store using an RFID portal. It was a logical move for “the supply chain” company. No other retailer bought into the concept.
4) The law of physics and the untidiness of Walmart’s own systems got in the way.
5) Walmart’s CIO was sent away, and the VC and analysts had to wait for mobility and social networking to come along.
That’s my summation.
In follow-on comments I’ll say RFID on cases and pallets were a solution looking for a problem. The magic was always going to be at item level in more expensive products than cans of peas.
Rather than further the adoption of RFID, I think the Walmart mandate set the industry BACK about 5 years. That’s why we’re starting to talk about it again now. Ultimately, it promises to do for certain sectors in stores what bar-code did for distribution centers–replace physical inventories with cycle counts.
RFID may not be moving in CPG, but nearly every apparel company has had a successful pilot. IT is now waiting for the business units to connect the dots and realize RFID’s potential to improve out of stocks and facilitate Web-to-store commerce. Don’t count it out yet.
The author stated, “CPG companies and Walmart did not do enough “hard math” to really understand the returns before making substantial investments in RFID.” I disagree. First of all, this isn’t “hard math.” With just one simple fact in this very article, “tags are now 5-7 cents,” I could tell you there isn’t an ROI business case for 95% of FMCG products. The other challenges are technology and true industry collaboration. If different types of packaging are difficult for the RFID readers to read the tags, there is a significant problem with real-world application. Further, as long as a typical supermarket carries items from hundreds of suppliers, if more than one method of accounting for the product is required (RFID and UPC, etc.), then the whole flow of goods, and supply chain visibility goal is compromised.
It’s just like having a small supplier that doesn’t have UPCs on their products. That simply is unacceptable today. Well, until RFID is embraced by the far majority of suppliers, the idea is only that…an idea.
In order to make item level RFID work in CG, the ROI is dependent on both trading partners utilizing the technology and the applications. It does not work unless all your suppliers (for retailers) or all of your distribution chain (for manufacturers) play.
RFID was over promised, pure and simple. The concept and principles are sound. The cost never approached payback for low-cost food items. Technology problems are still available to be solved. Basically, all or nothing works best in physical distribution. Having some items with RFID and others without does not work well or solve the problem. What will work is when a complete department, meaning all suppliers, uses something. For example if every HBC vendor put RFID tags on, retailers could work with it. Unless every item in the store has an RFID tag, the system only works through store receiving. In reality, RFID is in competition with existing security approaches for higher value merchandise. The problem is RFID needs food volume to get the cost down.
As this article intimates, there are a lot of reasons for the delayed results of RFID in CPG. In some sense, they parallel the challenges of electronic shelf tags. In both instances, the concept seems pretty straight forward and the benefits intuitive. Unfortunately, in both cases a great idea has been shot down by the reality of costs and technical capability.
Effective implementation of RFID requires not only purchasing the technology (tags and readers) but also requires changes in business processes at both manufacturers and retailers. I think these were especially underestimated for the transition phase while products had to be handled by both traditional and RFID based processes. There were not enough benefits identified for the initial investments required by manufacturers. (A very big one that I feel was underestimated is the ability to “own” the distribution channels by tracking serialized containers.) The manufacturer had the initial costs without corresponding benefits. Retailers do not have large R&D budgets, so they need quick paybacks. The classic “chicken and egg” situation evolved as they waited for manufacturers to put RFID tags on products. For those who can remember, retailers implemented UPCs (GTINS) as they labeled products in-store instead of price marking. The benefits of scanning were immediate and the marking process was changed, but was already in place.
The technical requirements of RFID turned out to be a lot more challenging than were expected (another similarity with electronic shelf tags). Success of RFID requires consistent implementation according to agreed upon standards. The challenge has been trying to develop standards with insufficient real world experience. Standards have had to evolve as new challenges are discovered. But the technology continues to improve and will eventually reach its initial expectations. The big question I see here is whether other technologies such as cheap CCD based cameras will challenge the need for RFID as it becomes possible to monitor the CPG supply chain with ubiquitous video.
The chief lesson from the RFID experience is one that has been taught many times when dealing with new technology. New technology often meets unexpected challenges, costs more, and takes longer to implement than expected. RFID makes sense now for asset tracking and situations (such as apparel) where there is a proliferation of product variations and exposure to shrink, or where counterfeiting (such as pharmaceuticals) is a significant issue. As the technology gets cheaper and better, it will also make sense for CPG.
As is usual when one of our pieces is used here, I feel inclined to make a few responses to some of these comments:
1. No one is “counting RFID out.” As stated, it is in fact apparently thriving in apparel–and soon shoes–right now, with a much clearer value prop, a more logical deployment plan, and importantly much better technology and cost based on the hundreds of millions invested for the original CPG focus. Many have said, and I agree, that apparel would not be doing what it is without that technology investment for the CPG/Walmart applications.
2. Don’t think the recession really had anything to do with the collapse of the CPG program, it was nearly dead before that, other than perhaps the recession putting the final nail in the coffin.
3. I have debated whether Walmart actually set the industry back, and have generally argued Yes, as does Paula. But that is with two maybes: (1) Yes compared to if Walmart had used a more rational approach (say starting with promotional execution and high ROI items; or (2) if other retailers would have begun piloting and experimenting, instead of waiting to see how Walmart turned out. But the lack of current effort out of even Germany’s Metro stores, which was more aggressive and bullish than Walmart, says something.
4. It was in fact Simon Ellis of Unilever who said the right ROI analysis (hard math) was not performed. If there “isn’t an ROI business case for 95% of FMCG products,” as one wrote above, than a lot of CPG companies were sure wasting a lot of money investing in the technology that couldn’t deliver–millions of dollars. Maybe it was “easy math” not hard, but the point is the investments were driven by a sort of euphoria that the value would manifest itself and I believe a reluctance for anyone inside CPG companies to “cry wolf” on the math for several years because you would be viewed as a barrier to RFID.
It’s time to dispel the myth, once and for all, that item-level RFID makes sense for moderately priced and expensive items but not for cheap merchandise.
These days, in the apparel category, millions and millions of items with retail price points in the $5-7 range are receiving disposable RFID tags.
Responding to Marshall’s comment on low value items, this is correct, which was implied in our full article, but I believe results from:
1. High turn as a substitute if you will for high value. Could be high margin as well.
2. The benefits of tagging the full category in the store, versus trying to mix RFID and traditional products in one category. There is research showing the extra benefit from full category tagging. Naturally then, if the category contains lower priced products, they would be tagged.
I was working for a supplier as a computer programmer-analyst when Walmart first laid out its RFID mandate. From the get-go, I thought Walmart was rushing things. But I thought to myself that if anyone could push through an RFID mandate at this speed, it would be Walmart. Now, however, it looks like their initiative has foundered, and I think rightly so. It was a case of too much too soon, in my opinion.
My original thought back then had been that Walmart should have started with only the pallet level, simply to introduce the concept to its suppliers and to only incrementally add to the suppliers’ costs and investments in new hardware and software (and training and procedures). Once Walmart had the concept of RFID at the pallet level firmly established in the industry, and its suppliers very comfortable and accepting of pallet-level RFID, then it would be time to take the next incremental step, which would be at the case level. Once everyone was on board at the case level, and beginning to realize some actual ROI on the technology, then the final frontier would have been the individual retail item, although, depending on the kind of item, this might never have been realized.
I think that the technology still holds a great deal of promise, but I think that it should be implemented in steps at a slower pace, rather than rushing in to realize the RFID utopia that everyone was promising back then.
“jmstone” you are quite wise.
The VERY BIG THING you have to give Walmart credit for in the evolution of RFID is the requirement that the tag data include a class (item) identifier. The initial vision that came out of MIT had the data content being a random index to a central data base of object definitions. While everyone was dithering about how to accomplish this, Walmart said “no” and demanded that the item identification be part of the tag data. If this hadn’t been done we probably would probably still be waiting for the first implementation.