Skip to content
E-commerce product segmentation by labels

Building a Custom Label Strategy for Shopping Campaigns

· by Digitelia · 4 min read

custom_label_0 through custom_label_4 are five free-text attributes in your Google Shopping feed. They don’t show to shoppers. They exist purely for you to bucket products for campaign segmentation. Used well, they’re the highest-leverage feature in Shopping campaign management — the difference between optimizing a 1,000-SKU catalog as one undifferentiated mass and tiering it strategically for vastly better ROAS.

This guide is the operational playbook: what to put in each label, how to use them in campaign structure, and the patterns that produce measurable ROAS lift.

Product segmentation strategy

Why custom labels matter

The default approach: one Shopping or Performance Max campaign for all products. Google’s algorithm optimizes across the entire catalog, picking which products to push based on conversion likelihood and value.

The problem: this optimizes for blended ROAS. Your high-margin best-sellers get the same tROAS target as your low-margin clearance items. The algorithm doesn’t know about margin; only revenue.

Result: clearance and low-margin SKUs absorb spend that should have gone to high-margin products. Your reported ROAS looks fine, but contribution margin (the real economic measure) underperforms what it could.

Custom labels let you split products into segments and assign different bidding strategies to each. Same conversion data, but the algorithm now optimizes per segment based on the goal you set.

The five labels — strategic uses

You have five labels. Use them for:

custom_label_0 = margin tier

The single most important label. Categorize products by margin contribution:

  • “high” (35%+ gross margin)
  • “mid” (20-35%)
  • “low” (under 20%)
  • “loss-leader” (negative or break-even)

custom_label_1 = product category at a level Google’s taxonomy doesn’t capture

Things like:

  • “best-sellers”
  • “new-arrivals”
  • “clearance”
  • “evergreen”

This is broader than Google’s product category — it’s your strategic bucket.

custom_label_2 = seasonality

  • “spring”
  • “summer”
  • “fall”
  • “winter”
  • “year-round”

Useful for seasonal campaigns and dayparting.

custom_label_3 = price tier

  • “under-25”
  • “25-100”
  • “100-500”
  • “premium” (500+)

Different price tiers often need different bid strategies and creative.

custom_label_4 = inventory status

  • “in-stock”
  • “low-stock” (under 20 units)
  • “drop-ship” (3rd party fulfillment)
  • “pre-order”

Useful for adjusting spend based on stock availability.

These are recommendations, not commandments. Pick the dimensions that matter most for your business. If margin and inventory are most important: use label 0 and 1 for those, leave others empty.

Campaign structure leveraging custom labels

Once labels are populated, structure campaigns to leverage them.

Pattern 1: Margin-tier campaigns

Most common, highest ROI:

Campaign A: PMax — High Margin

  • Filter: custom_label_0 = “high”
  • Bidding: tROAS 5.0-6.0
  • Aggressive — these products can absorb higher CPA

Campaign B: PMax — Mid Margin

  • Filter: custom_label_0 = “mid”
  • Bidding: tROAS 3.5-4.5
  • Moderate

Campaign C: PMax — Low Margin / Clearance

  • Filter: custom_label_0 = “low” OR custom_label_1 = “clearance”
  • Bidding: tROAS 2.0-2.5
  • Loose — goal is volume to move inventory

Same budget pool optimized to maximize blended contribution margin, not just blended ROAS.

Pattern 2: Best-sellers concentration

Identify top 20% of products that drive 80% of revenue. Tag with custom_label_1 = "best-seller". Run a dedicated campaign with aggressive tROAS.

The algorithm naturally favors converting products, but isolating them ensures budget stays concentrated there during volatile periods.

Pattern 3: Seasonal campaigns

Tag products with custom_label_2 = "summer" for swimsuits, beach gear, etc. Pause those campaigns in winter; run them aggressively in season.

Pattern 4: Inventory-aware bidding

Low-stock SKUs (under 20 units) shouldn’t get aggressive spend — you’ll sell out fast. Tag with custom_label_4 = "low-stock" and either pause or reduce bids on those.

For drop-ship items, often different margin profile — separate campaign with appropriate tROAS.

Custom labels in feed management

Populating custom labels at scale

Manual entry is impractical for 1,000+ SKU catalogs. Three automated approaches:

Approach 1: Merchant Center feed rules

Merchant Center has built-in feed rules that can populate custom labels based on other feed data.

Example rule: “If price > 100, set custom_label_3 to ‘premium’.”

Pros: managed in MC interface. No code. Cons: limited logic (no margin data unless explicitly in feed).

Approach 2: Backend / ETL pipeline

Your e-commerce platform exports product data including margins (from cost-of-goods database) and inventory. ETL pipeline transforms and adds custom labels before feeding to Merchant Center.

Pros: full data access. Can incorporate margin from accounting system. Cons: requires data engineering setup.

Approach 3: Feed management tools (DataFeedWatch, GoDataFeed, Channable)

Mid-ground. These tools sit between your platform and Merchant Center, allowing rule-based label population using broader data than MC alone.

Pros: easier than custom pipeline. Cons: monthly subscription.

For most accounts, start with Merchant Center feed rules. Graduate to feed management tools or ETL when complexity exceeds MC’s capability.

When labels change

Products move between tiers:

  • Clearance products that sell out → no longer clearance
  • New arrivals that prove themselves → become best-sellers
  • High-margin items that get discounted → become mid or low

Your label assignments need to update as these shifts happen. Frequency:

  • Inventory labels: daily (high turnover)
  • Best-seller labels: weekly (volume shifts gradually)
  • Margin labels: monthly (cost changes are slow)
  • Seasonal labels: quarterly

If labels stay frozen, campaigns optimize against stale data and performance degrades.

Reading reports leveraging labels

In Google Ads → Reports → custom report with custom_label dimensions.

Useful views:

Performance by custom_label_0 (margin tier): revenue, ROAS, contribution margin, CPC, conversion rate per tier. Surfaces which tiers are over- or under-funded.

Performance by custom_label_1 (strategic category): best-sellers vs. evergreen vs. clearance contribution.

Cross-tier analysis: best-sellers within high-margin vs. best-sellers within low-margin.

These reports support better budget allocation decisions monthly.

Common custom label mistakes

1. Not using labels at all. Massive missed lever. Even basic margin-tier segmentation typically delivers 15-30% ROAS improvement.

2. Using all five labels for unrelated data. Most accounts only need 2-3 labels with real strategic dimensions.

3. Static labels. Setting once and never updating. Stale labels degrade campaign performance.

4. Inconsistent label values. “high” in some products, “High” in others, “h” elsewhere. Standardize.

5. Labels with too many unique values. 50 unique values in custom_label_2 = no segmentation. Use 3-7 values per label maximum.

6. Labels not aligned with campaign structure. If your campaigns don’t filter by labels, the labels do nothing. Build campaign structure first; label to support it.

7. Ignoring inventory-based labels for thin inventory. Pushing PMax hard on a SKU with 5 units left = wasted spend (sold out before spend amortizes).

A 30-day custom labels implementation

Days 1-5: Strategy.

  • Define what each of your 5 labels will represent.
  • Map products to label values.
  • Confirm data sources for each (margin from accounting, inventory from ERP, etc.).

Days 6-15: Implementation.

  • Set up Merchant Center feed rules or pipeline to populate labels.
  • Validate label coverage (no missing values; consistent spelling).
  • Submit refreshed feed.

Days 16-22: Campaign restructure.

  • Split existing PMax/Shopping campaigns by label-based filters.
  • Set appropriate bidding strategy per tier.
  • Configure budget split.

Days 23-30: Measure and tune.

  • Compare new structure performance vs. prior baseline.
  • Adjust tROAS targets per tier based on data.
  • Document the label scheme for future maintenance.

Expected impact: 15-40% blended ROAS improvement after 60-90 days.

Frequently asked questions

Do custom labels affect SEO or organic Shopping placements? No — custom labels are internal to your account. They don’t appear in product cards or affect organic listing visibility.

Can I change a custom label without disapproving the product? Yes — labels are free-text and don’t have approval consequences.

How often should I update labels? Inventory: daily. Best-seller status: weekly. Margin tier: monthly. Seasonal: quarterly.

Should I use custom labels in Performance Max? Yes — PMax campaigns can be filtered by custom label values, allowing the same tiering strategy.

What’s the most impactful single label? custom_label_0 = margin tier. Single biggest ROI lever in Shopping campaign management.

Custom labels are operational segmentation, not glamorous strategy. The accounts that use them well consistently outperform accounts that don’t, even with identical creative, audiences, and bidding. The 30-day implementation pays back within the first 60-90 days of new campaign structure. If you’re not already using labels strategically, that’s the highest-leverage Shopping improvement available to you this quarter.

Tagged

#custom-labels#shopping#segmentation#google-ads#ecommerce#feed-strategy