Case Study: Home Decor Retailer Improves Performance 2.5x with Shopping Feed Optimisation
How feed optimisation and margin-based bidding transformed 8,000 SKUs into profitable advertising
ROAS vs Previous
Revenue Increase
Impression Share Increase
The Client
Industry: Home accessories and decor
Location: United Kingdom
Product Range: 8,000+ SKUs including furniture, lighting, textiles, kitchenware (£15-£500 price range)
Average Order Value: £85
Previous Google Ads Experience: Running Shopping campaigns with agency management
Monthly Ad Spend: £22,000
The Challenge
Primary Goal: Improve ROAS and scale revenue profitably across a large, diverse product catalogue with varying margins and performance characteristics.
Specific Pain Points
1. Poor Product Feed Quality
The client’s product feed had fundamental issues that limited performance:
- Generic product titles: “Item 12345 Blue” instead of “Velvet Blue Cushion Cover 45x45cm”
- Missing descriptions: 60% of products had no description field
- Inconsistent categorisation: Products in wrong Google categories
- Poor image quality: Small images, wrong aspect ratios, white backgrounds only
Impact: Impression share stuck at 35% (Google didn’t understand what products were).
2. Treating All Products Equally
Previous agency ran one “all products” Shopping campaign with a single ROAS target (350%). This ignored critical product differences:
- High-margin lighting: 70% margins, fast sellers
- Medium-margin textiles: 45% margins, steady demand
- Low-margin furniture: 18% margins, competitive pricing
Impact: Budget wasted on low-margin furniture while high-margin lighting was underfunded.
3. Massive Negative Keyword Gaps
With 8,000 products, search queries were matching inappropriately:
- Cushions showing for “cushion foundation” (makeup)
- Table lamps showing for “table salt lamps” (wellness products)
- Mirrors showing for “wing mirrors” (car parts)
Impact: 15-20% of budget wasted on completely irrelevant traffic.
4. No Product Performance Visibility
The client couldn’t see which products drove profit vs which wasted budget. Reporting showed overall metrics, not product-level detail.
Impact: Unable to make data-driven inventory or bidding decisions.
The Strategy
Phase 1: Product Feed Overhaul (Weeks 1-3)
Before touching campaigns, I completely rebuilt the product feed:
📝 Product Title Optimisation
Priority 1: Top 500 products (40% of revenue) manually optimised
Example transformations:
- ❌ Before: “Item 45678 Grey”
- ✅ After: “Luxury Velvet Cushion Cover 45x45cm Grey Soft Furnishings”
- ❌ Before: “Lamp Black Table”
- ✅ After: “Modern Black Table Lamp Adjustable Metal Desk Light”
Formula: [Material] + [Colour] + [Product Type] + [Size/Dimensions] + [Key Feature] + [Category]
Remaining 7,500 products: Bulk template applied with automated keyword insertion
📄 Description Enhancement
Created rich descriptions for all products using template system:
- Material composition and quality indicators
- Dimensions and sizing information
- Care instructions and usage context
- Style and design benefits
- Complementary products and room suggestions
Result: 100% product coverage (vs 40% before)
🏷️ Custom Labels for Margin-Based Bidding
Added custom label structure to enable strategic bidding:
- custom_label_0: Margin tier (high/medium/low)
- custom_label_1: Product category (lighting/textiles/furniture/kitchenware)
- custom_label_2: Performance tier (bestseller/standard/new/clearance)
- custom_label_3: Price band (£0-£30/£30-£100/£100-£250/£250+)
Purpose: Enable campaign segmentation and granular bidding control
Phase 2: Campaign Restructure (Week 4)
Replaced single campaign with 6 strategically segmented campaigns:
✓ Campaign 1: High-Margin Lighting (70% margin)
ROAS Target: 320% (volume focus)
Budget: 35% of total (£7,700/month)
Strategy: Aggressive scaling, bid to dominate
✓ Campaign 2: High-Margin Textiles (65% margin)
ROAS Target: 350%
Budget: 25% of total (£5,500/month)
Strategy: Steady growth, seasonal scaling
✓ Campaign 3: Medium-Margin Kitchenware (45% margin)
ROAS Target: 420%
Budget: 20% of total (£4,400/month)
Strategy: Balanced profitability
✓ Campaign 4: Medium-Margin Decor (40% margin)
ROAS Target: 450%
Budget: 10% of total (£2,200/month)
Strategy: Selective bidding on winners
✓ Campaign 5: Low-Margin Furniture (18% margin)
ROAS Target: 650%
Budget: 5% of total (£1,100/month)
Strategy: Strict control, brand awareness only
✓ Campaign 6: New Product Launch
ROAS Target: 300% (flexible)
Budget: 5% of total (£1,100/month)
Strategy: Test new products, graduate winners to main campaigns
Phase 3: Negative Keyword Buildout (Weeks 4-8)
Systematic search query analysis to eliminate wasted spend:
🚫 Irrelevant Category Exclusions
Added 2,000+ negative keywords across campaigns:
- Makeup: foundation, blush, bronzer, mascara (cushions showing for wrong “cushion”)
- Automotive: wing mirror, car mirror, headlight (mirrors/lamps matching wrongly)
- Wellness: salt lamp, himalayan lamp, therapy lamp (table lamps matching)
- Clothing: hat stand, coat rack (furniture matching apparel searches)
- DIY: paint, drill, hammer, screw (tools, not home decor)
Result: 22% reduction in wasted clicks on irrelevant queries
Phase 4: Competitive Analysis (Ongoing)
Monthly auction insights review to understand competitive landscape:
- Identified 8 major competitors (Dunelm, Next, John Lewis, Wayfair, etc.)
- Tracked impression share overlap and position above rate
- Monitored competitor pricing changes and promotional activity
- Adjusted bidding strategy based on competitive intensity by category
The Results
12-Week Performance (Post-Optimisation)
Above Industry Avg
ROAS Improvement
Revenue Growth
💰 Revenue Impact
- 55% increase in total revenue (£342k vs £221k over 12 weeks)
- Budget held constant at £22k/month (efficiency, not spend increase)
- £121k additional revenue from same ad spend
- 32% improvement in profit margin (revenue flowing to high-margin products)
📊 Efficiency Metrics
- Impression share: 68% (vs 35% before) – 94% increase
- Click-through rate: 1.8% (vs 1.1% before)
- Conversion rate: 4.2% (vs 2.9% before)
- Average order value: £98 (vs £85 before) – better product mix
🎯 Campaign-Level Performance
| Campaign | Performance vs Target | Revenue Share | Strategy Result |
|---|---|---|---|
| High-Margin Lighting | +52% above target | 42% | ✅ Scaled to 42% (from 35%) |
| High-Margin Textiles | +26% above target | 28% | ✅ Outperformed expectations |
| Medium-Margin Kitchenware | -4% under target | 18% | ✅ Still highly profitable |
| Medium-Margin Decor | On target | 8% | ✅ Meeting expectations |
| Low-Margin Furniture | +18% above target | 2% | ✅ Profitable at low volume |
| New Product Launch | +13% above target | 2% | ✅ Testing successfully |
📦 Product-Level Insights
- Top 100 products: Now drive 60% of revenue (vs 40% before)
- Budget concentration: Winners get 3x more spend than losers (vs equal before)
- New product success rate: 65% of new products achieve target ROAS within 30 days
- Long-tail contribution: Bottom 6,000 SKUs drive 15% of revenue profitably
Key Success Factors
1. Feed Quality as Foundation
Optimising product titles increased impression share from 35% to 68%—a 94% improvement. Google’s algorithm needs quality data to match products to searches. Poor feed = invisible products, regardless of bid amounts.
2. Margin-Based Campaign Segmentation
Separating products by margin profile was transformational. High-margin lighting could bid aggressively (320% ROAS target, achieved 485%). Low-margin furniture bid conservatively (650% ROAS target, achieved 520%). One ROAS target across all products would have either over-spent on furniture or under-funded lighting.
3. Systematic Negative Keyword Management
2,000+ negative keywords eliminated 22% of wasted spend. Large catalogues naturally match irrelevant queries—proactive exclusion is critical, not optional.
4. Custom Labels for Control
Adding margin tier, category, performance tier, and price band labels enabled precise segmentation without manual product selection. The feed structure drove campaign strategy.
5. Product Performance Transparency
Monthly product-level reporting showed the client exactly which items drove profit. This informed inventory decisions, pricing strategy, and new product development—Google Ads became a business intelligence tool, not just advertising.
Client Testimonial
“Our previous agency never mentioned feed quality—they just said our performance was ‘industry standard.’ Peter’s first recommendation was completely rebuilding our product feed. It seemed like busywork at the time, but the results proved him right. Impression share doubled, ROAS improved 2.5x above industry average, and revenue increased 55% at the same budget. What really impressed us was the product-level reporting. For the first time, we could see which products were profitable advertising channels vs which wasted money. That data changed our entire inventory strategy. Peter didn’t just improve our Google Ads—he gave us business intelligence that drives decisions across the company.”
— E-commerce Director, Home Accessories Retailer
Ongoing Optimisation (Months 4-12)
Continuous Feed Improvements
- Quarterly title audits: Review top 1,000 products for keyword opportunities
- Seasonal description updates: “Perfect Christmas gift” messaging for Q4
- New product integration: 50-100 new SKUs monthly added with optimised data
- Image quality upgrades: Lifestyle images replacing plain product shots (15% CTR increase)
Campaign Refinements
- Budget reallocation: High-margin lighting scaled from 35% to 45% of budget
- New campaign added: “Bestsellers” campaign for proven winners (significantly exceeded targets)
- Seasonal strategy: Christmas campaign November-December (exceeded target, £68k revenue)
- Competitive adjustments: Increased bids during competitor sale periods to maintain visibility
Year 1 Total Results
Annual Revenue
Above Previous Year
Revenue Increase
Lessons Learned
✅ What Worked
- Feed quality first: Doubled impression share before touching campaigns
- Margin-based segmentation: Budget flows to profitable products automatically
- Custom labels: Enabled precise control without manual product selection
- Negative keyword discipline: Weekly reviews prevented waste accumulation
- Product-level reporting: Data informed inventory and business strategy
⚠️ What Didn’t Work Initially
- Over-segmentation: Initially tried 12 campaigns; 6 was optimal (too many = learning phase issues)
- Conservative new product ROAS: 300% target was too relaxed; 350% worked better
- Ignoring long-tail: Bottom 6,000 SKUs contributed 15% revenue—can’t ignore them
Why This Case Study Matters
Large catalogue retailers face unique challenges:
- Managing thousands of products across Shopping campaigns
- Varying margins requiring different profitability thresholds
- Feed quality issues limiting impression share
- Irrelevant query matching wasting budget
- Lack of product-level performance visibility
This case study demonstrates that feed optimisation and strategic segmentation deliver better results than simply increasing ad spend. The client achieved 55% revenue growth at constant budget—pure efficiency improvement.
Is Your Product Feed Holding Back Performance?
If you have a large product catalogue and suspect your feed quality is limiting impression share or wasting budget on irrelevant traffic, let’s talk. I’ll audit your feed and show you exactly what’s possible with proper optimisation.
✓ 20+ Years Shopping Campaign Experience
✓ Product Feed Optimisation Specialist
✓ Large Catalogue Expert
✓ Consistent Above-Industry Performance