With Google Shopping and product-feed-based online advertising amazingly growing year over year, both in spend and return on investment, it is essential for advertisers to stay up-to-date with the latest changes in order to always get the most out of this extremely valuable channel.
Google released many new shopping features in the past 8 months. Supplemental feeds were launched in September last year, feed rules got a massive update around February this year and the brand new "goal-optimized" shopping campaigns (also known as "universal shopping campaigns") and shopping actions were just launched last month. All these new shiny features wouldn't be helpful at all without a smart use of the so-called custom labels.
Let's dive in what custom labels are and how to get the most out of them:
- Start here if you are a complete Google Shopping Beginner
- What custom labels are?
- What custom labels are used for?
- How to set up custom labels for Google Shopping
- Custom labels' usage examples
- Preferred Custom Labels by Google Shopping Experts
Start here if you are a complete Google Shopping Beginner
Product feeds can come in multiple formats and forms. XML, CSV, TXT, etc.
Sometimes, understanding how the feed attributes and fields work is not really straightforward. Therefore, for the purpose of this guide let's think of a product feed as a simple spreadsheet showing the list of products you have in your catalogue.
Rows show the list of products and columns the products' attributes, like product ID, Title, Description, Price, Availability and so on. Let's assume you would place the ID attribute in the first column of the sheet, column A. Then you place the title in column B, the description in column C and so on.
In Google Shopping, some columns (also called attributes, fields or product data) are mandatory, while others are optional. (Click here for more details about mandatory and optional attributes)
Product Feed Example:
What custom labels are?
Google Shopping enables you to create product groups based on Category, product Type, ID, Condition, Brand, Channel, and Channel exclusivity.
If you would rather break down your product groups by another attribute, custom Labels on Google Shopping are offering you the opportunity to create groups of products based on conditions you are able to set yourself.
Custom Labels are 5 optional columns, named custom_label_0, custom_label_1, custom_label_2, etc.., that can be added to a product feed. The values in the custom label columns are at the discretion of the advertisers. In fact, there are no requirements or rules dictated by Google.
Custom Labels allow for a greater freedom when it comes to shopping bidding strategies. Adding these values for every individual product allows you to segment the biddings on your Product Listing Ad’s within your Shopping Campaigns.
While the other feed attributes are tied to a specific product and always have to be the same as shown on the advertiser's website, custom labels can be anything and potentially change all the times you want.
In fact, their value is not showing anywhere else than in your AdWords account.
What custom labels are used for?
Let's imagine the following common scenario: You are the PPC manager for a medium/large e-commerce business. You have hundreds of thousands of products in your catalogue. They are definitely too many to be managed manually through Excel or Google Sheet. Therefore, you ask your developers to generate a product feed for you.
Unfortunately, developers are always super-busy doing something else and can't spend the right amount of time looking after your requests. Therefore, in order to make their job as simple and quick as possible, you would ask them to add only the mandatory basic attributes to the feed.
Then, all you would need is just a small favour.
You need them to add 5 extra empty columns, the infamous custom labels. Through these magical empty columns, you will be able to add additional fields by yourself, without asking the developers to edit the feed on their end, which would usually take a long time and many emails back and forth. You could add a special sale price, a sale start and end date, the product category, an additional description, shipping costs and much more.
So, I would say that the main use of custom labels is to allow marketers to be as independent as possible from developers. If you are a brand, this is particularly helpful when the developer team is outsourced to an agency.
Or, if you are an agency, custom labels allow you to work on the feed without requesting changes from the client, which makes your work quicker and smoother. Having that said, let's go a bit deeper and see what are the main custom labels' functions when it comes to Google Shopping campaigns.
1. Product segmentation and shopping campaigns optimization
One of the main complaints Google Shopping campaigns have been receiving is the lack of control you can have on them. In fact, unlike regular search campaigns, you can't bid on keywords.
Therefore, you are losing control over when product listing ads are showing and, most importantly, over their cost per click.
Google shows your product ads trying to match the search query to the title and description of your products in the feed. But when you have hundreds of thousands of products, the chances a product comes up for irrelevant searches are very high. This is also due to the fact that many of them are likely to share the same description or have similar titles. This is particularly true, for example, for different pieces in the same furniture collection, or different books by the same author. All you want to do then is getting into the driver's seat and gaining full control over your shopping campaigns.
The best way to achieve control is by creating product groups in AdWords. With product groups, you can group together the products you want and manage them separately. For example, if you are a fashion retailer you might want to set a higher bid for shirts than for shoes. Or maybe you want to show a kind of shoes only in a specific region. Maybe you want some trousers to be visible only from mobile. The possibilities are endless, but what you ultimately want to do is segmenting the product feed in order to manage different product segments in different ways.
Now, the question is: how can I group together only the products I want?
Back to the PPC-manager-vs-developers example, if you are lucky enough to be a close friend with the best developer in the company, you could ask him to add a product_type column to your feed and sort your product groups based on this attribute. Or maybe you can ask him to add a custom optional column called 'product_segment', where he aggregates the products based on your specific requirements. However, if the best developer in the company doesn't even know your name, chances are you would need to sort things by yourself.
And that is where custom labels come in. You can add values to the custom label columns directly from your Google Merchant Center account, without changing or editing the actual feed. The values will be visible in AdWords straight after their creation and from there you will be able to create product groups based on your custom labels.
A proper product segmentation also gives you the ability to access product-group-level reports. In fact, once you have created multiple product groups, you will be able to see their performance at a glance.
Product-level reporting, although available in AdWords under the 'products' tab, is definitely not a viable option for large e-commerce businesses with huge feeds. Rather, reports at the product group level are much more insightful and easier to read.
3. Run experiments
Through product-group-level reporting, you are able to test a segment performance against another one.
For example, you might want to see how the shipping cost influences the conversion rate. So you would create a product group aggregating all products which are shipped at a X price and another group of products shipped at a Y price. Then you can compare the performances. How would you be able to segment products this way? Easy, populating a custom label column based on the products' shipping price.
How To Set Up Custom Labels For Google Shopping
Let's see how you can set Custom Labels directly in the feed.
Create Custom Labels Using Static Value in DataFeedWatchWith the 'static value' rule you can extrapolate a word from any product attribute and populate a custom label column with that same word. In the example below, the rule looks up all products where the word 'shirt' appears in the title and populate a custom label column with the same word. This way I can aggregate all shirts into one product group in AdWords.
Create a Custom Label for Season or other attributes for which you do not have a separate field:
Create a Custom Label for Margin based on price:
Create a Custom Label for Top Sellers. You can label individual products as a Top Seller by mapping it from the product ID:
If you have a list of Topsellers, you can also just upload them to DFW and use 'is-in-list':
Create Custom Labels with Rename
With the 'Rename' rule, you can in fact set a custom label to a value which already exists either in your primary or supplemental feed. One of the most common usages of this rule would be populating the GTIN column with the product ID values (if the two are equal). This way, developers don't need to create the GTIN column themselves and you can save them additional work. But it is when coupled with the supplemental feed feature that the 'set to' rule gives its best.
Create a Custom Label for Color or size, or weight or age or gender or any attribute that is already in a field in your feed:
Custom Labels' Usage Examples
Custom Labels can be used for many purposes. How to best take advantage of them really depends on your specific business' needs. Here are just a few scenarios where custom labels can be essential to improve a shopping campaign performance and return on ad spend. Take these examples as an inspiration and be creative in finding new ways they can help your business.
1. Bid more aggressively on bestselling and seasonal products
When it comes to large e-commerce websites, chances are only a relatively small fraction of the product catalogue is actually relevant at a given time. For example, for a bookshop, the bulk of sales would be mainly driven by the most trending books of the season. Given your time is limited and you can't optimize all the thousands of products you have in the feed, it makes sense to prioritize just those books. What you can do then is pulling a report for the bestselling products of the week and putting them into a Google Sheet. Then you can add a column called, for example, 'segment' and populate it with the value 'Bestsellers'. As long as the product IDs you have entered in the Google Sheet match the ones in your primary feed, you can then upload this sheet as a supplemental feed in Merchant Center. At that point, you can use the 'set to' rule to set custom_label_0 in your primary feed to be equal to the 'segment' attribute in your supplemental feed. See below how your supplemental feed might look like:
You will then find the custom label's values in your AdWords account. From there you will be able to create at least two product groups, one containing all the products with custom_label_0 equals 'Bestsellers' and the other with all the rest of the catalogue.
Furthermore, using Google Sheets formulas you can dynamically change the custom labels' values. For example, as soon as a product is not a top seller anymore, the value can automatically change from 'Bestsellers' to 'Low Priority'. You can also split product segments in different campaigns, rather than just product groups. This way you can allocate different budgets rather than just bids. This is particularly helpful when you use automated bid strategies like Target ROAS or Maximize Clicks, which don't allow for manual bids.
2. Be more competitive on high-value/high-margin products
You can use the same strategy (supplemental feeds + custom labels) to segment products based on their strategic value, which can be either about margin or price. Again, you can either bid more aggressively on valuable products or put them in a separate campaign with a higher budget. You can also create multiple segments based just on the product price. For example, you can aggregate all products priced under €5, and then between 5 and 15, between 15 and 30 and so on. Even though you are still not sure how to bid on them, segments can still be useful to just see how those products perform as an aggregate. Based on that, you will be able to set optimal bids down the line.
3. Be less competitive on products with missing or wrong attributes
Sometimes it can happen that the feed generated by the developers is not 100% accurate. Some products might be missing the description, which is highly recommended but not mandatory. Or maybe some others just have a default image because the actual one wasn't able to be sourced. Large feeds are full of inaccuracies like those. Although you can use supplemental feeds to fix some of these issues, chances are you won't be able to fix all of them. Therefore, you want to exclude, or at least set a low bid, on products with missing attributes, as they basically look bad compared to the competitors' ones. You can go to the diagnostic tab in your Google Merchant Center account and download the full list of products with attribute issues. Then, you can add them to a supplemental feed and create a column which you will use to populate a custom label attribute in your primary feed. This can be done through the aforementioned 'set to' feed rule. The value you can use in this case might be something like 'missing attribute'.
Before you can start using custom labels you have to think about how you want to organize them, of course. Using a table like the one below helps you defining the right objectives and values that fit your specific needs.
Based on the examples before this could look like this:
|0||Blockbusters||High turnover, Low turnover|
|1||Budget Burners||High revenue, Low revenue|
|2||Poking||New, Clearance, Sale|
|3||Profit||High margin, Low margin|
|4||Season||Christmas, Superbowl, Winter|
4 Google Shopping experts share their preferred Custom Labels
One of my favorite custom labels for Google shopping is <Product name>.
It may sound weird, but in the Google interface you can subdivide products by Brand, Product type, channel, Item ID etc. etc. - but you can’t subdivide by product titles or product name. See screenshot 1.
So here is a simple trick to subdivide by product names (titles) in the Google interface by using custom labels and feed rules.
To be able to do it in the first place, you’ll need to make a tiny tweak in Merchant Center where you set Custom label 0 to be equal to title by using Feed rules. See screenshot 2. By doing this, we are assigning Custom label 0 a value from the existing feed – in this case product title. After this, you can go to the Google AdWords interface and subdivide by Custom label 0 – and just like that you have your product names in the view instead of for example Item ID.
Limitations: Be aware that there is a limitation to custom labels and you can only have 1.000 unique values – so this should work just fine in accounts with less than 1000 products.
Products subdivided by Item ID vs. Products subdivided by Product Title.
Another favorite custom label is price buckets
Price bucket segmentation or grouping is classic and a great way to separate products by price range. This is also useful when there are large variations in product prices. By grouping by price buckets, you will easily get an overview of which product range products that provide you with the most value for your campaign and company.
For example, a product that costs $150 could be in the price range ‘$50-$150’.
You can define several price buckets. For example:
< $50 $50 - $150 > $150
You can create price bucket custom label directly in Merchant Center by using feed rules or use any other third party feed optimizations tools.
What we normally do, is that we use feed rules where we set a condition if price is greater than ‘XX’$ then set Custom label to “ProductPriceOverXX”
Last but not least we have competitive pricing – custom label
If your pricing on some of your products is competitive then why not take full advantage of it? For some clients, we are using the competitor's price in Custom labels. This way we can see if our client’s products are cheaper or more expensive. By knowing this, we can bid more aggressively on the products that we now know are cheaper than the ones of the competitors – assuming it will perform better ☺.
There are several tools out there which can be used to get the competition prices and display inside the Custom label.
Beginners. A simple way to configure custom labels is to base it on product pricing, while product pricing often is a good indicator for earnings.
When using the custom label for product pricing I usually arrange groups of low, medium and high product prices: 0-500DKK, 501-1.500DKK and 1.501DKK to max price in the feed.
Set campaign priority to low, medium and high and bids in that order.
Experts. Besides arranging your products according to product price, you can use the custom label for creating the optimal ‘3 campaign solution’; Generic, Brand and Model/Product. Set the campaign priority as high, medium and low respectively and bids in reverse order.
Next add all model and brand terms as negatives to the generic and all models/products to the brand campaign. I prefer using a shared budget for all three of the campaigns, while this allows you to bid lowest for the generic, more for brand and the most for model/products.
Using the custom label is also possible for seasonal inventory and if you have more than e.g. 8 products in stock of a product and want them sold, you can automatically place them in a higher priority campaign.
Custom labels are first and foremost about bid differentiation across Shopping campaigns and groups of products.
Identifying key differences in expected product performance lets you adjust your bids accordingly. Below are some of my favorites that can be set up by using DataFeedWatch.
1. Sales vs. normal price items: Divide your Shopping campaigns into two tracks to enable higher bids on sale items, as you can expect higher CVR and you may want to clear out stock of sales items first.
2. Gross margin brackets: If you have this information available, you can create custom labels with intervals of either absolute gross margin levels, or as a percentage of revenue. This means that you can divide Shopping campaigns into different CPA targets (if absolute) or ROAS targets (if ratio), depending on the profitability.
3. Bestsellers (of the season): Creating a label for your best selling items, possibly also of the current season, is a very powerful strategy to bid up on the products that are most likely to convert right now.
Depending on the industry, you may need to update this quite frequently (e.g. for books). It may not be the PPC specialist who is best informed to know what will be a bestseller in advance, so basing the labels on a “is in list”-rule, means that another person with this information can update it ongoing from a shared spreadsheet.
One of my favorite labels is definitely the “On Sale”-label. We often see that our clients are more price competitive compared to the competitors, when having products on sale, therefore we tend to raise the focus on those products. Price is also a ranking factor on Google Shopping, and it is important not to miss out on conversions due to, for example, low bids.
Some other of my favorite labels are our season labels, which we use according to the seasons: Black Friday, Christmas Sale, January Sale, High Season/Low Season, etc.
This way, it is much easier to adjust the bids higher and lower according to the season – for example, we want to bid more for swimwear, when we approach the summer compared to what we want to in the middle of winter. Besides there are always better options in selling in the high season when bidding higher on items that you may not sell that much of normally. You may say that we use Custom Labels to switch between the different seasons in a much easier and efficient way.
Since "Data-driven" is in our DNA, we always work with as much data as possible. During the years we have helped many clients, including some of Denmark's largest companies. When we work in our clients' Adwords accounts we never do anything, which is not Data-driven. We do deep analysis of all the data we can find, and we work actively with our clients' data. One of the things we work with is labels in Adwords campaigns, and custom labels for Google Shopping campaigns.
Two of the labels we love the most, and most often use are "Season" and "Margin". These labels are incredibly helpful, when we work with bidding strategies. If you don't use labels, and just want to start out with one label, then go for "Margin". This is a need to have.
Why are the labels "Season" and "Margin" so useful?
These labels are useful because you can define values such as "Winter", "Spring", "Summer", "Fall" (for Season) or "LowMargin", "HighMargin" (for Margin) to better target your customers with the right products at the right time, and have control over your bids. If the demand for a product doesn't change during the year, you don't need "Season". If you run AdWords campaigns try to check out your Dimensions data first. Does demand change during the year? If the answer is "yes", then you should use "Season" for your Google Shopping campaigns. To understand if a product is "LowMargin", "HighMargin", we work closely with our clients, as they usually have this data. This information is very valuable to us, so we can determine the best bidding strategy, so that we not only get as many conversions as possible, but also achieve the highest profitability for our clients.
Conclusion: Why does all of this actually matter?
Google is pushing more and more its automated bidding solutions and smart campaigns driven by machine learning. Goal-Optimized Shopping Campaigns are just the latest example of this. Therefore, you might ask yourself if all of this segmentation actually still matters.
Well, the answer is yes, it still does, more than ever before. With product listing ads getting a higher and higher click share on the Google result page, it is of primary importance for advertisers to show their products in the highest ranking as possible in order to defeat competitors and get the click through.
In fact, the product listing ads section on Google can show up to 30 products on desktop (less on mobile), but just a bunch of them are visible without 'swiping'. You want to be among the top ones, as obviously they are the most likely to be clicked.
At a recent Google retail conference in Dublin, it was said that 45% of PLAs' clicks are on the very first result. You want your products to be that result. Given that you can not bid on keywords and you cannot decide when and where to show specific products, the only way to effectively making sure you show your best products in the best ranking as possible is by segmenting your feed through product groups and custom labels. Also, multiple product groups give more signals to the Google machine learning system, the same way multiple audience lists do.
Therefore, you want to segment as much as possible even if you are already taking advantage of Google automated bid strategies or universal shopping campaigns.
Download our Guide to Data Feed Optimization eBook to learn more about:
- How to improve the quality of your product feed?
- Different shopping channel business models and which one works best for you?
- What tactics to use to improve the shopping campaign ROI.
Recommended Further ReadingThe Hidden Feature of Custom Labels - shows how product segmentation can be helpful to find new strategies and test business hypothesis.