This is the second in a series of blog posts looking at the parallels between locations and products.
In our first post on location and product information, we looked at the history of both street addresses and latitude/longitude and their parallels to product information. We ended by saying that the Global Trade Item Number (GTIN) isn’t the Prime Meridian for commerce. In other words, while GTIN is a great start to global standardization, it’s not going to be the final way we navigate the product universe.
Let’s back up a bit, however, and talk about what the GTIN actually is. GS1, the standards organization behind UPCs and GTIN, says that GTIN can be used to identify types of products at any packaging level and unambiguously identify trade items online. In other words, every product at every size and pack should have a unique GTIN — a unique address of sorts. This GTIN can be encoded with barcodes or RFID tags to efficiently track and process products. By its description, GTIN sure sounds like the Prime Meridian, right? Unfortunately, some aspects of GTIN implementation better mirror the era of the pendulum clock and others mirror the era of multiple meridians.
In the pendulum clock era, we knew what longitude was, but didn’t have a good way to measure it at sea without an accurate timepiece. In the era of multiple meridians, we knew what longitude was and finally had an accurate way of measuring it. But we didn’t have a common point of reference. In 1879, 95.5 percent of the world’s ships used one of 11 different main meridians, leading to major confusion.
Because GTINs are a superset of other standards (namely ISBNs, UPCs and EANs), they fall victim to any downfalls inherent in these standards. While ISBNs are strictly regulated and easy to understand, UPCs (which are a subset of EANs) do not share that advantage. There are four main drawbacks to UPCs, two of which fall into the pendulum clock problem and the other two of which fall into the multiple meridians problem.
Companies that need to use UPCs must buy a company prefix, which can get expensive. A company with up to 10 products requiring UPCs will need to pay $250 for the initial company prefix fee and an additional annual renewal fee of $50. A fee of $25 per product will substantially increase setup costs for new organizations. Looking at it in terms of production costs, the least expensive and highest volume tier costs between $1.05 and 10.5 cents per product when the company buys the prefix and between 21 cents and 2.1 cents every year thereafter.
UPCs aren’t unreliable in and of themselves any more than a pendulum clock is unreliable on land. UPCs in the current Pervasive Commerce environment, however, act more like pendulum clocks at sea. Two primary issues cause this unreliability: disreputable UPC barcode vendors and retailers who do not enforce valid UPCs.
The expense of buying a company prefix from GS1 has prompted an entire industry around reselling UPCs. Honest companies who do this buy a company prefix for 100,000 products and then resell subsets of that group at a markup that’s still significantly less expensive than buying the smallest tier from GS1. These companies aren’t causing the issue, as long as they maintain their annual renewal fees from GS1. Unfortunately, there are other companies that simply fabricate UPCs and sell them. These fake UPCs pollute the environment, which makes a UPC sometimes untrustworthy, like that pendulum clock.
Some retailers exacerbate the pollution by allowing fake UPCs into their environment. Others have documented how the UPCs that Amazon requires and uses in its marketplace practically encourage faking. At Indix, we’ve seen direct evidence of retailer UPC pollution: when we gather data from retail sites, we often see UPCs tagged that have letters in them—real UPCs only have numbers.
The pendulum-clock-like unreliability of UPCs causes a huge headache to begin with, but combining unreliability with the way valid UPCs get assigned creates even more of a problem.
When a company buys a company prefix from GS1, the organization provides guidelines of how to assign them. Unfortunately, companies interpret and implement the guidelines differently, meaning that they don’t conform to a single standard. For example, one company might assign a UPC to every different color of a shirt but not to every different size, but other might assign a different UPC to each color and size combination.
Perhaps the most confusing multiple-meridian problem with GTINs is that they can be reused. Section 188.8.131.52 of the GS1 General Specification states:
A GTIN allocated to a trade item that has become obsolete must not be re-used for another trade item until at least 48 months have elapsed after:
■ the expiration date of the last original trade items produced with that number
■ the last original trade items produced with that number have been supplied to the customer.
In the case of clothing, the minimum retention period is reduced to 30 months.
While GS1 advises that brand owners should consider waiting longer for certain products, the current re-use guidelines completely discount the thriving marketplace economy. With the prevalence of used products being sold online, a single GTIN could easily refer to multiple products. By definition then, GTINs cannot ever be unique identification numbers.
The way that GS1 talks about GTINs sounds like they’re gearing up for the product equivalent of January 1, 1914, when the longitude system and the adoption of the current Prime Meridian were implemented globally. However, how can we get to global adoption when we’re still measuring with pendulum clocks and trying to measure from multiple meridians? Without GTIN accuracy and standardization, we can’t.
GTINs could be almost there, however. If GTINs were regulated more strictly like ISBNs, they would be far more reliable and much better to use as the global product identification standard. ISBNs do a few things differently:
GTINs aside, there are a lot of other ways that we talk about products that make me think of how we talk about locations. In my next post, I’ll talk further about naming and addresses used for products.
Also published on Medium.