It’s a bit more nuanced than that
When May 16 arrives, life is going to be a pain for retailers wanting to list on Google’s shopping platform.
So life probably has already been a hell for you, especially if you’re the sorry soul in charge of finding GTINs for your listed products.
Some of the retailers we’ve spoken to mentioned that they plan to get around this by marking all items which lack GTINs as
‘identifier exists’ = false,
in their feed.
Google will eventually flag these products, and you’ll be caught with your pants down.
So what happens if you don’t have GTINs when you need to have one?
You lose to your competitors that are more prepared, and worse, you lose out on precious traffic.
- February 8, 2016: Warnings begin. You’ll start to see item-level warnings in the Diagnostics tab for products that don’t meet the requirements. Use these warnings to help you update your product data.
- May 16, 2016: Enforcement begins. You’ll start to see item-level disapprovals in the Diagnostics tab for products that don’t meet the requirements. After this date, you’ll need to meet the GTIN requirements to continue serving ads for your products.
Once enforcement begins, you can expect to receive red flags for even products that are supposed to have GTINs, but were marked as unnecessary in your feeds.
Why, Google, why?
Well, from our sources, we learned that these requirements were driven by a couple of core reasons:
- Filter out fake merchants and fake product listings
- Improve the credibility and trust in the Google Shopping platform
- Improve price comparison and product matching, through better product mapping using unique identifiers
These items ultimately help Google in offering a better product for their users.
A more credible and attractive platform for shopping / e-commerce helps drive user traffic to the platform, and helps retailers sell more.
My products are getting flagged. Help!
Our sources clarified Google’s policy in further detail — much more than what we were hearing from retailers desperate to pack their catalogs with literally any GTIN just to get their products through verification:
Google is adopting a “blacklist” strategy as opposed to a “whitelist” approach to flagging products that need GTINs.
They are aware that not all products have GTINs.
In order to have a GTIN, the product’s manufacturer must:
- Register with GS1 to obtain a manufacturer’s prefix
- Issue a GTIN for the product itself.
In fact, not all products manufactured as even issued GTINs.
Dyson, for example, issues GTINs for its suite of products, but not for individual components or spare parts.
This is because GTINs are not cheap. Each manufacturer’s prefix has a limited number of permutations that can be created, based on the length of the prefix.
Generally, the shorter the prefix, the greater the number of GTINs you can create for your products. But more doesn’t always mean better. Getting a GTIN prefix isnt cheap either as you can see below:
This state of the matters has led to 2 things:
- Manufacturers aren’t incentivized to issue GTINs for all of their products, just a select few.
- The parties who actually benefit from GTINs (e.g. retailers and logistics companies) are not allowed to create GTINs for their products, meaning which they have to rely on manufacturers to ensure GTINs are being created.
This sorry state of affairs is well known to Google, which is why they only flag products that have a reasonable chance of having a GTIN issued (i.e. has a manufacturer’s prefix, and is a complete product, not a component).
This actually is great for retailers, because you don’t have to worry about looking for GTINs that don’t exist (which is what is happening right now).
And we can help you with that!
Here’s how we can help with your GTIN-pocalypse
At Semantics3 we have been working closely with retailers, resellers and other players in the e-commerce space who rely on Google Shopping and Product Listing Ads to drive traffic and sales.
In order to find GTINs for your product, there must first exist a GTIN.
Many products don’t have GTINs. It’s often unnecessary to find GTINs for these products. Examples of products that usually don’t need GTINs include clothing, fashion accessories and spare parts/components.
On the other hand, products that often require GTINs include items like electronics and home appliances. Keep in mind, these products that have GTINs must be manufactured by companies that have obtained a GS1 prefix, which you can obtain here.
Once all these requisites are satisfied, we will be able to locate the GTINs for your products.
One way to filter out these products is to simply wait until Google flags them for you — this would allow Google to do the hard work of figuring out which products need GTINs (which dramatically improves the rate for GTIN-matching).
Typically we would need some form of metadata to ingest in order to generate matches. Data fields that work really well include:
- Retailer URLs
- Model Part Numbers (MPN)
- Amazon ASINs
- Other unique identifiers like ISBNs, EANs, etc.
While we can find a GTIN based on the product name or brand, this approach is not advised since text-based searches are fundamentally fuzzy matches and have a degree of error involved.
A typical process of GTIN-matching involves the following:
- Clean and validate the input data, including checking URL and SKU validity
- Ingest the data into our product matching algorithms to extract unique identifiers or URLs
- Test the unique identifiers against the database to find matches
- If matches are not found, use additional steps like importing data from other retailer catalogs, deploy search crawlers to locate product records and so on.
The process is often multi-pass, using several matching techniques to find the matches, which is why we almost always deploy our stellar Datascience-as-a-Service team to assist you with a high-touch service.
Semantics3 will be attending GS1 Connect 2016 in Washington, D.C. June 1-3 — Come find out how we can help you at the #GTIN-pocalypse Survival Stand (Booth 42).
Lovingly built in San Francisco, Singapore, and Bengaluru by Hari Viswanathan and the Semantics3 Team.