Custom Label Intelligence: Segment, Bid, and Measure Granularity In Google Shopping

Granular performance tracking in Google Shopping and Performance Max lives or dies on how you use five small fields: custom_label_0 to custom_label_4.

Used well, they become a control panel for:

• Segmenting products by performance, margin, price position, and stock
• Sending the right SKUs into the right campaigns
• Reading results at label and SKU level, not just at campaign level

Used badly—or not at all—and you’re stuck with blended ROAS, manual spreadsheets, and guesses about where budget is leaking.

If you’ve been searching for tools for segmenting Google Shopping campaigns and tracking performance by custom labels, or evaluating Google Shopping optimization platforms with advanced segmentation and granular tracking, this guide is for you.

This article walks through a practical system for:

• Designing a custom label schema that matches real bidding decisions
• Implementing labels 0–4 in Merchant Center and Google Ads
• Building dashboards that track performance by label and down to SKU
• Running a simple experimentation ladder on top of that structure
• Seeing how Product Metrics automates the heavy lifting

1. Custom Labels 0–4 The Switchboard For Granular Segmentation

Google Shopping and Performance Max give you five custom label attributes in the product feed: custom_label_0 through custom_label_4. Each product can hold one value per label, and those values are entirely defined by you. (Source: Google Ads Help – Use custom labels for Shopping ads)

Google’s own docs confirm that:

• You can have up to five custom labels per product, numbered 0–4
• Each label holds one value per product (for example, custom_label_0 = stars)
• Labels are used to subdivide product groups and filters in Shopping and Performance Max campaigns, and for monitoring, reporting, and bidding on those groups
• Each label allows up to 1,000 unique values per account, with basic length limits per value

(Source: Google Ads Help – Custom labels 0–4 attribute)

From Google’s perspective, custom labels are deliberately generic: they’re just text fields. What makes them powerful is the meaning you attach to them and the way you wire that meaning into campaigns and reporting.

1.1 Treat Labels As Owned Decision Dimensions

Because you only get five of them, custom labels should be treated as owned decision dimensions, not as a dumping ground for miscellaneous tags.

A good rule of thumb:

• Give each label a single, stable meaning across the account
• Only create values that you’re prepared to act on with different bids, budgets, or priorities
• Prefer performance and economics (how products behave in ads) over static merchandising trivia

Product Metrics’ own glossary definition matches this: custom labels are defined as five feed attributes (custom_label_0–4) used to group products in campaigns, with one value per label per product, and they’re explicitly recommended for things like performance segments and stock status. (Source: Product Metrics – Custom labels glossary)

2. A Practical Custom Label Schema For Shopping And PMax

Here is a concrete schema that works well for most ecommerce catalogs and maps cleanly to campaign decisions.

You don’t have to adopt these exact names, but the shape—performance first, economics second, context third—is what matters.

2.1 Suggested Schema

custom_label_0 – Performance Segment
Purpose: capture how a product behaves in ads, based on click volume and return vs your ROAS/POAS target.

Example values (aligned with Product Metrics’ segmentation):

• stars – high volume, above target return
• question_marks – high volume, just under target
• cash_cows – low volume, above target
• dogs – high volume, below target
• drainers – low volume, no return
• ghosts – not serving (no impressions)

Product Metrics uses machine learning to place every product into one of these six segments based on advertising clicks and ROAS or POAS, then writes the segment into a Merchant Center custom label via a supplemental feed. (Source: Product Metrics – Product Segmentation)

custom_label_1 – Margin Or Profit Band
Purpose: distinguish between products that generate very different profit per sale.

Example values:

• high_margin, mid_margin, low_margin, negative_margin

With cost-of-goods data, segmenting by margin tiers inside a custom label is a standard recommendation for making bids profit-aware rather than revenue-only. (Source: FeedArc – Custom labels from GA4 performance)

custom_label_2 – Price Position Vs Market
Purpose: overlay how competitively priced a product is.

Example values:

• below_median, at_median, above_median

Product Metrics’ Competitor Prices feature, for example, compares your products to competitors, computes a price position, and can write that into Merchant Center as a custom label (e.g. custom_label_2 = above_median). (Source: Product Metrics – Product-Level Google Shopping Optimisation)

custom_label_3 – Stock Lifecycle State
Purpose: capture whether the product is in a special state that should change how aggressively you advertise it.

Example values:

• new, evergreen, clearance, low_stock, end_of_line

custom_label_4 – Experiment Cohort Flag
Purpose: track A/B tests or cohorts without rebuilding campaigns.

Example values:

• test_feed_titles_v1, test_pricing_v2, control

You can also follow Google’s own suggestion to use one label for seasonality, one for margin, etc.; the key is to document the schema and avoid repurposing labels later. (Source: Google Ads Help – Use custom labels for Shopping ads)

3. Implementing Labels 0–4 In Merchant Center And Google Ads

Once you’ve designed the schema, you need a clean way to get it into your feed and into campaigns.

3.1 Populate Labels In The Feed

There are three common routes:

1. Directly in the primary feed
   • Your ecommerce platform or feed tool writes custom_label_0–4 into the main feed that Merchant Center ingests.
2. Via a supplemental feed
   • You upload a second feed keyed on product ID that only contains custom labels; Merchant Center merges it onto the primary product data. Google’s documentation and Product Metrics’ glossary both describe this pattern: the primary feed holds core product data, while a supplemental feed supplies segmentation labels such as custom_label_0 = stars. (Source: Google Merchant Center Help – Supplemental feeds; Product Metrics – Supplemental feed label glossary)
3. Through an optimization platform
   • Tools like Product Metrics connect to Google Ads, Merchant Center, and your store, compute segments and business rules, and write approved custom labels back to Merchant Center via API on a daily schedule. (Source: Product Metrics – Google Shopping monitoring)

Whichever route you use, follow Google’s technical constraints: up to five labels per product, one value per label, length and uniqueness limits, and a 24–48 hour propagation time for labels to appear in Google Ads. (Source: Google Ads Help – Use custom labels for Shopping ads)

3.2 Wire Labels Into Shopping And Performance Max

In Standard Shopping:

• Use inventory filters at the campaign level to include only certain label values (e.g., custom_label_0 in (stars, cash_cows) for a “high-priority winners” campaign)
• Within campaigns, subdivide listing groups (product groups) by custom_label_0–4 to see performance by segment and adjust bids at that level

In Performance Max:

• Use labels in listing group filters inside asset groups, so that each asset group can target, for example, only stars and cash_cows
• Because custom labels are exposed as subdivision dimensions for listing groups, they can drive separate budgets and targets for each segment. (Source: FeedArc – Custom labels from GA4 performance; Product Metrics – Custom labels glossary)

The structure doesn’t have to be complex. A simple three - campaign setup based on custom_label_0—Stars & Cash Cows; Question Marks; Dogs & Drainers—already creates very different levers for budget and bid strategies.

4. Dashboard Recipes For Label Level And SKU Level Tracking

Good segmentation is wasted without clear, label-aware reporting. Out of the box, Google Ads exposes product-level and product-group reports, and its API offers detailed views such as shopping_performance_view and product_group_view that can be sliced by product attributes. (Source: Google Ads API – Product reporting for Shopping ads)

Here are dashboard “recipes” you can build in a BI tool or in a specialized platform like Product Metrics.

4.1 Segment Health Performance By custom_label_0

Objective: see how spend, revenue, and profit split across performance segments.

Dimensions

• custom_label_0 (performance segment)

Metrics

• Spend, clicks, conversions, revenue, ROAS
• If you have cost-of-goods: gross profit, POAS (profit on ad spend)
• Product count per segment
• Share of total spend vs share of total revenue/profit

Questions to answer weekly:

• Are Stars and Cash Cows taking the majority of spend and revenue?
• Are Drainers and Dogs consuming more spend than their share of revenue or profit justifies?
• How many Ghosts (no impressions) do you have, and are they worth fixing?

Product Metrics’ own product overview shows this kind of table out of the box: spend, revenue, ROAS, and the share of total revenue and spend per segment (Stars, Question Marks, Cash Cows, Dogs, Drainers, Ghosts). (Source: Product Metrics – Product-Level Google Shopping Optimisation)

4.2 Economics Overlay custom_label_0 × custom_label_1

Objective: combine performance with margin tiers to prioritize where to push or pull back.

Dimensions

• custom_label_0 (performance segment)
• custom_label_1 (margin band)

Metrics

• Revenue, gross profit, POAS
• Impression share, potential uplift if you increased impression share on high-margin winners

This reveals, for example:

• High-margin Stars: candidates for more aggressive targets and budgets
• Low-margin Stars: revenue-heavy but profit-thin; maybe cap budgets or increase prices
• High-margin Question Marks: worth feed and landing-page work to turn into future Stars

4.3 Price And Competitiveness custom_label_0 × custom_label_2

Objective: understand how price position interacts with performance.

If you tag products with a price position label (e.g. below_median, at_median, above_median), your dashboard can show:

• How many Drainers are also above_median in price
• Whether Stars tend to sit at market price or below

Product Metrics’ Competitor Prices module is built exactly to create this context and send price position into Merchant Center as a custom label (e.g., custom_label_2 = above_median), so you can report and act on it directly. (Source: Product Metrics – Product-Level Google Shopping Optimisation)

4.4 SKU Level Deep Dives Within Each Label

Objective: move from segment-level insight to SKU-level actions.

For each performance segment, create a view that lists:

• Top products by wasted spend (high spend, low revenue or POAS)
• Top products by incremental potential (high POAS, low impression share)
• Products that changed segment in the last sync

With Google Ads’ Shopping Performance View you can retrieve impressions, clicks, cost, and conversions by product item ID, then join that to your label values and product metadata. (Source: Google Ads API – Shopping Performance View reference)

Product Metrics automates this by letting you segment by custom_label_0 and then drill down to products, journeys, competitor prices, and inventory—all tied back to each SKU. (Source: Product Metrics – Product Segmentation)

5. An Experimentation Ladder For Label Led Optimization

Once labels and dashboards are in place, you can climb a simple experimentation ladder rather than trying to redesign everything at once.

Step 1 Instrumentation Only No Structural Changes Yet

• Ship custom_label_0–2 into Merchant Center
• Build the segment health and economics dashboards above
• Run for 2–4 weeks to establish a baseline; remember new labels can take up to 48 hours to appear fully in Google Ads reporting. (Source: Google Ads Help – Use custom labels for Shopping ads)

Outcome: you know how spend and outcomes currently split across segments, before any campaign rebuilds.

Step 2 Simple Segment Based Campaigns

Now start restructuring campaigns around labels—but keep it simple:

• Campaign A – Stars & Cash Cows
  Filter: custom_label_0 in (stars, cash_cows)
  Strategy: higher budgets, slightly looser target ROAS/POAS to buy incremental volume

• Campaign B – Question Marks
  Filter: custom_label_0 = question_marks
  Strategy: budgets for testing; focus on feed and pricing fixes to see if they can graduate into Stars

• Campaign C – Dogs & Drainers
  Filter: custom_label_0 in (dogs, drainers)
  Strategy: very conservative targets, lower priority, or even exclusion

A Product Metrics campaign-structure guide lays out similar mappings: performance segments written into custom_label_0, then campaigns filtered on those labels with different bid strategies per segment. (Source: Product Metrics – ML Product Segmentation for Google Shopping)

Run this structure alongside your old one with reduced budgets until you’re confident in the new setup.

Step 3 Add Margin And Price Overlays

Once the performance-only structure is stable, layer in economics:

• Use custom_label_1 to give high-margin Stars more aggressive targets than low-margin Stars
• Use custom_label_2 to treat above-median-price Dogs as pause candidates, while below-median Dogs might be rescued with title or image improvements instead

Step 4 Feed Level Experiments

Finally, use labels to run clean experiments on feed and pricing changes:

• Assign a subset of products to a test cohort via custom_label_4
• Change titles, images, or price only for that cohort
• Compare performance against a control group with the same baseline labels

Feed-level A/B tests need time: many practitioners recommend at least 2–3 weeks of clean data per variant to separate noise from real uplift. (Source: FeedArc – Measuring the ROAS impact of a product feed change)

6. Why Dedicated Tools Beat Spreadsheets For Granular Tracking

You can build all of this in spreadsheets and generic BI tools, but at a cost:

• Manual label updates: every change to performance segments or price position risks breaking your own rules
• Fragmented data: Google Ads holds product-level metrics; GA4 holds journeys and multi-product orders; your platform holds margin and stock
• Experiment fatigue: it’s hard to keep an experiment log tied to label values and campaign changes over time

When you’re specifically looking for Google Shopping campaign segmentation tools with customizable performance tracking granularity, or campaign segmentation solutions with detailed performance dashboards, the real question is:

How much of the segmentation, labeling, and label-level reporting does the platform maintain for you, and how much are you still hand-stitching?

A strong platform should:

1. Operate at SKU level (not just campaign or ad group)
2. Provide consistent custom label governance (clear meanings per label 0–4)
3. Write labels back to Merchant Center automatically and safely
4. Offer built-in dashboards by label and SKU, with profit if you supply costs
5. Make it easy to run and interpret experiments on top of the structure

7. How Product Metrics Gives You Custom Label Intelligence By Default

Product Metrics is built specifically around this workflow: segment products by performance and economics, turn those segments into custom labels, and close the loop with Shopping and PMax campaigns.

7.1 Segment First Six ML Driven Performance Groups Per SKU

Once connected, Product Metrics uses machine learning to place each product into one of six segments—Stars, Question Marks, Cash Cows, Dogs, Drainers, Ghosts—based on ad clicks and return vs your ROAS or POAS target. (Source: Product Metrics – Product-Level Google Shopping Optimisation)

Those segments are visible in its own dashboards (Segment → Product → Context → Action), so you can see at a glance how much spend and revenue each group accounts for, and which segments are soaking up budget without paying back. (Source: Product Metrics – Google Shopping monitoring)

7.2 Turn Segments Into Merchant Center Labels Automatically

On Pro plans, Product Metrics’ Product Rules and Merchant Center labels integration:

• Map performance segments and business rules (e.g., POAS ≥ 1.4, margin ≥ 40%, stock > 8 weeks) to label values such as custom_label_0 = stars or star_candidate
• Write those labels into Merchant Center via a daily supplemental feed, so campaigns can target them without brittle manual feed rules
• Keep labels fresh as performance and context change, while you approve the mappings

Product Metrics’ own documentation describes this “from segment to campaign” flow: segments become labels in Merchant Center, updated daily, and campaigns are then split by those label values. (Source: Product Metrics – ML Product Segmentation for Google Shopping; Product Metrics – Pricing)

7.3 Add Margin Stock And Price Context To Labels

Beyond performance segments, Product Metrics can:

• Pull in margin data from your ecommerce platform so you can create margin bands in another label slot
• Compute price position against competitors and expose it as a custom label (e.g., custom_label_2 = above_median)
• Use Inventory Insights to surface slow-moving or out-of-stock items so you aren’t paying for clicks you can’t fulfill

These context layers feed into Product Rules, which in turn write Merchant Center labels your Shopping and PMax campaigns can filter and subdivide on. (Source: Product Metrics – Product-Level Google Shopping Optimisation; Product Metrics – Pricing)

7.4 Label Aware Dashboards And Reporting

Because Product Metrics reads Google Ads and Merchant Center directly, its dashboards natively understand:

• Performance by segment (Stars vs Drainers, etc.)
• Performance within segments at SKU level
• How that performance changes as labels and Product Rules evolve

This makes it effectively a Google Shopping campaign segmentation tool with detailed performance dashboards, but based on your custom label schema rather than a fixed black-box view. (Source: Product Metrics – Tool-by-use-case: Shopping/PMax reporting software)

7.5 Pricing And Getting Started

Product Metrics is priced as a graduated share of your rolling 30-day Google Shopping spend, with the first €10,000 per month free and marginal rates from 0.65% down to 0.05% above that, with no minimum fee. (Source: Product Metrics – Pricing)

Show the dataHide the data

Marginal Product Metrics rate per band of rolling 30-day Google Shopping spend

Rolling 30-day Google Shopping spend	Marginal rate (excl. VAT)
€0–10K	0.00%
€10–25K	0.65%
€25–100K	0.60%
€100–200K	0.50%
€200–500K	0.35%
€500K–1M	0.27%
€1–2M	0.22%
€2–5M	0.15%
€5–10M	0.10%
€10M+	0.05%

There are three low-friction ways to start:

1. Free analysis (no credit card)

   • Log in with Google at app.productmetrics.io
   • Connect your Shopping account and see your own products segmented into six groups, on your data
   • Feed export with Merchant Center labels is not included at this step, so it’s safe for evaluation
     (Source: Product Metrics – Pricing; Product Metrics – Book a demo)

2. 30-day Pro trial

   • Start the Pro plan trial when you’re ready for label export and deeper reporting
   • Includes feed export with Merchant Center labels, Product Journeys, Competitor Prices, Inventory Insights, and Full Signal Tracking
     (Source: Product Metrics – Pricing)

3. Founder-led demo

   • Book a call with founder Berend Vrakking at productmetrics.io/demo
   • Walk through segmentation, context, and label export on demo data, then get a demo account for your team
     (Source: Product Metrics – Book a demo)

If you’re evaluating platforms that offer SKU-level performance tracking for Google Shopping campaign segmentation, or looking to buy software for campaign segmentation with high-granularity performance tracking, the core question is whether the tool can:

• Maintain a clear, opinionated custom label schema for you
• Keep those labels in sync with real performance and economics
• Give you label-aware dashboards and experimentation paths that stay stable over time

Product Metrics is built to answer “yes” on all three.

When you’re ready to move beyond manual slices of Google Ads and fragile feed rules, connect your accounts, let the ML segmentation run, and start managing Google Shopping at the level where decisions are actually made: custom labels and SKUs.