Resolving Product Overlap in Performance Max Campaigns to Boost Efficiency for a Leading E-commerce Brand
Industry: E-commerce Retail (Electronics, Apparel, Home Decor)
Ad Spend: $1M+/month on Google Ads
Campaign Type: Performance Max (PMax)
Region: North America
—
🔍 Background
Shoporia (made up name as I cannot disclose their name due to NDA) is a fast-growing e-commerce marketplace managing a vast catalog of millions of SKUs across multiple product categories. The internal digital team had divided responsibilities across business verticals, each running their own Performance Max campaigns to meet growth targets.
Despite their scale and investment, the Google Ads performance was plateauing:
Rising Cost Per Conversion (CPC)
Declining ROAS
Internal concerns about wasted spend
As part of my audit engagement, I was tasked with identifying inefficiencies in the structure and proposing a roadmap to enhance performance without increasing budgets.
—
🚨 The Problem
While reviewing the campaign structure and feed setup, I noticed a fundamental issue:
Multiple PMax campaigns were promoting the exact same products (Item IDs) at the same time, without coordination.
For example:
Campaign A (Electronics) and Campaign B (Home Essentials) both contained Item ID: XYZ123 – a Smart LED Bulb.
Since both campaigns were PMax, they were competing blindly in real-time auctions, unaware that they were bidding against each other.
Because PMax does not offer keyword-level or auction overlap data, the cannibalization was going undetected. The internal algorithmic bidding was being driven by multiple asset groups bidding for the same audience/product intent, resulting in:
Inflated CPCs
Confusing ad creatives being shown to users
Higher impression share loss
This scenario is known as “self-competition” — a silent budget killer.
—
🧠 Strategy & Diagnosis
To confirm the issue, I took the following steps:
- Data Export: Pulled all item IDs from every active PMax campaign (via Product Listing Group reports and asset group breakdowns).
- Conflict Detection:
Used Google Sheets + Google Apps Script to automate the process of identifying overlapping Item IDs across multiple campaigns.
Developed a script that compared item ID lists column by column, flagged duplicates, and grouped them by campaign.
- Campaign Ownership Mapping:
Created a matrix to define product ownership:
Which campaign had higher priority?
Where was the product performing better?
What was the product’s margin and strategic value?
Identified underperforming campaigns where SKUs were wasting impressions.
- Exclusion Plan:
Excluded overlapping SKUs from lower-priority campaigns.
Updated feeds using Custom Labels (e.g., label_1 = pmax_owner_electronics) for long-term automation.
Set up campaign rules in the Google Merchant Center feed to prevent future conflicts.
—
🛠 Implementation Steps
Reviewed Google Merchant Center feeds and removed duplicate feed entries.
Applied item exclusions at the Asset Group level within each PMax campaign.
Refined targeting and creative messaging to align with new product assignments.
Monitored performance closely via a custom Looker Studio dashboard focused on:
Conversion rate
CPC
ROAS
Impression share
Auction overlap (where available)
—
📈 Results (30 Days After Optimization)
📊 Metric 🔴 Before ✅ After 📉 Change
Cost Per Conversion $14.87 $9.32 ↓ 37.37%
Conversion Rate 3.1% 4.8% ↑ 54.8%
Impression Share Lost (ISL) 22.5% 9.3% ↓ 58.7%
Return on Ad Spend (ROAS) 3.8x 5.4x ↑ 42.1%
Wasted Impressions (estimated) 300K+/month <100K/month ↓ 66.7%
—
🎯 Key Learnings
PMax campaigns are not set-it-and-forget-it. Despite being automated, they require manual guardrails to avoid internal cannibalization.
In multi-team or multi-agency setups, lack of SKU visibility across teams can result in self-bidding, inflating auction costs.
Product ownership discipline is key. Each SKU must be mapped to a single campaign unless there’s a strong use-case (e.g., brand vs. generic campaigns).
Regular feed audits and label-based segmentation can go a long way in keeping PMax efficient.
—
🧰 Tools & Technologies Used
Google Ads (PMax Campaign Management)
Google Merchant Center (Feed Management)
Google Apps Script (custom duplicate-checking automation in Google Sheets)
Google Sheets for product mapping and reporting
Looker Studio for performance dashboards
Slack & Asana for cross-team coordination
—
📌 Final Thoughts
This case is a classic example of how automation without coordination can work against you. With a simple SKU audit powered by Apps Script automation, we resolved internal auction conflicts and realigned the campaigns — unlocking a 42% higher ROAS without increasing budget.
