Custom Labels and Product Segmentation
Custom labels are one of the most powerful yet underused features in Google Shopping. Five simple fields in your product feed can transform how you structure campaigns, allocate budgets, and apply bidding strategies. This guide covers the practical use cases, implementation methods, and campaign structures that make custom labels indispensable for serious Shopping advertisers.
The Five Custom Label Fields
Google's product data specification provides five custom label attributes: custom_label_0 through custom_label_4. Each field accepts a text value of up to 100 characters. You define what each label represents and populate it with values that are meaningful to your business.
Custom labels do not affect ad display, query matching, or auction behavior. They exist purely for campaign management. Their value comes from enabling product group subdivisions in Shopping campaigns and listing group segmentation in Performance Max, allowing you to apply different bids, budgets, and strategies to different product segments.
A common allocation of the five labels:
| Label | Purpose | Example Values |
|---|---|---|
| custom_label_0 | Margin tier | high_margin, medium_margin, low_margin |
| custom_label_1 | Performance bucket | hero, sidekick, villain, zombie |
| custom_label_2 | Seasonality | summer, winter, evergreen, holiday |
| custom_label_3 | Price range | under_25, 25_50, 50_100, over_100 |
| custom_label_4 | Product status | new_arrival, bestseller, clearance, standard |
There is no mandatory allocation. Use the labels in whatever way best serves your campaign management needs. The key is consistency: once you define a label's purpose, maintain that definition across your entire catalog.
Use Case: Margin Tiers
Segmenting products by gross margin is one of the highest-impact uses of custom labels. Different products generate different profits per sale, and your bidding strategy should reflect this.
Define margin tiers based on your cost of goods sold relative to the selling price:
- High margin (above 50%): These products can afford aggressive bidding. Higher CPCs are justified because each conversion generates substantial profit.
- Medium margin (25 to 50%): Standard bidding. These products should be profitable at normal CPC levels.
- Low margin (below 25%): Conservative bidding required. Each click needs to have a high likelihood of converting to maintain profitability.
With this segmentation in place, you can create separate product groups (or even separate campaigns) for each tier and apply different ROAS targets. High-margin products might target 300% ROAS, while low-margin products need 800% or higher. When you add the CSS advantage from Cobiro, every tier benefits from lower CPCs, making even low-margin products more viable for Shopping advertising.
Margin-based segmentation is most powerful when combined with the cost_of_goods_sold attribute in your feed. Providing COGS data enables Google Ads to display gross profit metrics in reporting, letting you see actual profit per product rather than just revenue. See Product Data and Attributes for more on implementing COGS data.
Use Case: Performance Buckets
The performance bucket approach, popularized by Producthero (now part of Channable) through their Labelizer tool, segments products based on historical Shopping performance. The framework categorizes products into four groups:
Heroes
Your top-performing products. High impressions, high clicks, high conversions. Heroes deserve the most aggressive bidding and the largest share of budget. They have proven they can convert profitably and should be given every opportunity to scale.
Sidekicks
Products with decent impressions and clicks but lower conversion rates than heroes. Sidekicks have potential but are not performing at the highest level. They may need title optimization, better images, or improved landing pages to convert better.
Villains
Products that consume budget without generating proportional returns. High impressions and clicks but low or no conversions. Villains need to be investigated: is the product page the problem? Is the price uncompetitive? Are the wrong queries triggering the ads? Consider reducing bids or excluding villains temporarily while diagnosing the issue.
Zombies
Products with very few or zero impressions. Zombies are invisible in the Shopping auction, receiving no traffic. The issue is typically poor feed data (weak titles, missing attributes), low bids, or low demand. Before writing off zombies, try improving their feed data. Many zombies can be revived with better titles and complete attributes.
The Producthero Labelizer automates this classification by analyzing historical performance data and assigning labels dynamically. Products move between buckets as their performance changes over time. The Labelizer tool works independently of your CSS choice, meaning you can use it alongside Cobiro CSS for maximum impact.
Update performance bucket labels regularly (weekly or bi-weekly) to reflect current performance. A product that was a hero last month may have become a villain due to increased competition or seasonal changes. Dynamic labeling ensures your bidding strategy stays aligned with actual performance.
Use Case: Seasonal Flags
Seasonal labeling helps you adjust bidding strategy based on demand patterns. Products can be tagged as:
- Summer / Winter / Spring / Autumn: Products with seasonal demand peaks
- Holiday: Products that sell primarily during holiday periods (Christmas, Black Friday, Easter)
- Evergreen: Products with consistent year-round demand
- Back-to-school / Valentine's / Mother's Day: Event-specific products
During a product's peak season, increase bids or move it to a higher-budget campaign. During off-season, reduce bids or exclude it entirely to avoid wasting budget on low-converting traffic. Seasonal labels make it easy to implement these shifts through product group bid adjustments rather than manual product-by-product changes.
Use Case: Price Ranges
Products at different price points attract different buyer behavior. Segmenting by price range allows you to tailor bidding accordingly:
- Low price (under 25 euros): High volume, lower AOV. Bid conservatively since each conversion generates less revenue.
- Mid-range (25 to 100 euros): Balanced volume and value. Standard bidding typically works well.
- Premium (over 100 euros): Lower volume, higher AOV. Can justify higher CPCs because each conversion is more valuable.
Price range labels are particularly useful when combined with margin labels. A high-margin, premium product is your ideal candidate for aggressive bidding. A low-margin, low-price product needs careful cost control.
Additional Use Cases
New Arrivals and Clearance
Label new products as "new_arrival" and clearance items as "clearance". New arrivals might receive a temporary bid boost to gain initial visibility and data. Clearance items might have a separate campaign with a lower ROAS target, since the goal is to move inventory rather than maximize profitability.
Bestseller Flagging
Mark your top 10 to 20% of products by revenue as "bestseller". These products have proven demand and typically convert well. Ensuring they have maximum visibility is often the highest-ROI bidding decision you can make.
Competitor Sensitivity
Label products where you are the exclusive seller differently from products sold by many competitors. Exclusive products can often achieve higher ROAS because there is less price comparison happening. Commoditized products need more competitive pricing and may benefit from different bidding targets.
LabelWiser from BigShopper
LabelWiser is BigShopper's automated labeling tool that enriches your product feed with data-driven custom labels. Unlike the Producthero Labelizer, which focuses on performance buckets, LabelWiser emphasizes competitive positioning. It analyzes your products against competitors in the market and labels based on factors like competitive price position, number of competing merchants, and price gap analysis.
This competitive intelligence can be mapped to custom labels and used for bidding decisions. Products where you are the cheapest seller may benefit from more aggressive bidding, while products where you are significantly more expensive might need lower bids or exclusion.
Implementation Methods
Supplemental Feed
The most common approach. Create a supplemental feed (Google Sheets, CSV, or XML) containing product IDs and custom label values. Upload it to Merchant Center and link it to your primary feed. This keeps label management separate from your primary feed generation, making it easy to update labels without touching your website or CMS. Supplemental feeds are the recommended approach for most merchants.
Feed Rules in Merchant Center
Merchant Center's feed rules allow you to set custom labels based on conditions in your product data. For example, you can create a rule that sets custom_label_3 to "premium" for all products with a price above 100 euros. Feed rules are useful for simple, attribute-based segmentation but lack the sophistication needed for performance-based labeling.
Primary Feed Integration
Some e-commerce platforms and feed management tools allow you to populate custom labels directly in the primary feed. Shopify apps, WooCommerce plugins, and feed tools like Channable and DataFeedWatch offer custom label functionality. This approach integrates labeling into your existing feed workflow, but updating labels requires regenerating the feed.
Cobiro Label Manager
Cobiro provides label management functionality that works alongside CSS to help you segment products effectively. The integration between CSS and label management means you can manage both cost optimization (through CSS) and campaign segmentation (through labels) from a single platform, simplifying your Shopping management workflow.
Custom label values are case-sensitive in Google Ads product group subdivisions. "High_Margin" and "high_margin" are treated as different values. Standardize your label values (all lowercase is the common convention) and ensure consistency across your feed to avoid creating unintended product group splits. Also note that custom label values have a maximum of 100 characters each.
Campaign Structure Using Custom Labels
Once your labels are in place, you can build campaign structures that leverage them. Here are two common approaches:
Single Campaign, Label-Based Product Groups
Keep all products in one Shopping campaign but subdivide product groups by custom label. For example, subdivide by custom_label_0 (margin tier), then within each margin tier, subdivide by brand or category. Apply different bids to each margin tier. This approach is simple and works well with Smart Bidding, which can optimize across all product groups within the campaign.
Multiple Campaigns, Label-Based Segmentation
Create separate campaigns for different custom label values. For example, one campaign for "hero" products and another for "sidekick" products, each with its own budget and bidding strategy. This gives you independent budget control per segment. Hero products get a larger budget with aggressive Target ROAS, while sidekicks get a smaller budget with a more conservative target.
Regardless of structure, the CSS advantage from Cobiro applies equally to all products in all campaigns. Custom labels control how you allocate budget and bids. CSS controls the cost efficiency of every bid. The two work together to maximize Shopping profitability.
Save 20% on Shopping Ads
Switch to Cobiro CSS in minutes. No disruption, no re-learning period.
Start your free trial