US-based auto parts store with 50,000+ SKUs across brakes, suspension, engine components, and aftermarket accessories. ROAS was stuck at 200–250% for over a year under a previous agency. The account wasn’t failing — it was stagnating.
The root problem was invisible: 25% of the catalog (12,500+ products) was disapproved in Google Merchant Center. Missing GTINs, incorrect brand mapping, and product titles with zero fitment data. A title like “Front Brake Pads” instead of “2019 Ford F-150 Front Brake Pads — Bosch OEM Quality.” For auto parts, fitment is everything — customers search by vehicle, and if titles don’t contain Year/Make/Model, Google can’t match the product.
One Performance Max campaign handled all 50,000 SKUs with one bidding target. A $5 air filter and a $400 turbocharger shared the same strategy. No segmentation by margin, vehicle popularity, or performance tier. Browser-only tracking was missing ~20% of purchases due to iOS restrictions and ad blockers.
Rewrote all 50K product titles using automated scripts pulling Year/Make/Model from the fitment database. Fixed disapprovals — missing GTINs, incorrect brand mapping. Approval rate went from 70% to 96%, recovering 13,000+ previously invisible products.
Created labels across three dimensions: margin tier (A: 30%+, B: 15-20%), vehicle popularity (F-150, Silverado = dedicated asset groups), and performance tier (top 500 SKUs flagged as hero products).
Hero Products (top 500 SKUs, aggressive bids), Catalog (long-tail, conservative), Brand Defense, and Clearance. Each tier with its own ROAS targets matching the economics of those products.
Implemented Server-Side GTM to recover the ~20% of conversions browser-side setup was missing. Algorithms finally had complete purchase data to optimize against.
First 3 months: spend stayed flat while efficiency gains compounded. Months 4-6: budget increased 20-30% per cycle while ROAS held above 400%.
Revenue grew while ad spend stayed flat for the first 3 months — pure efficiency gain from better feed matching, smarter segmentation, and complete conversion data.
Hero product tier (top 500 SKUs) delivered 65% of total revenue. Ford F-150 and Chevy Silverado parts — segmented into own asset groups — generated 35% of Shopping revenue from ~8% of catalog.
They found problems in our product feed that our previous agency never even looked at. Within two months, our ROAS nearly doubled — and we didn’t increase our ad budget.
How we doubled conversions from 3,400 to 6,490 and cut CPA from $2.73 to $1.56 for a gold buying business in Uzbekistan — by removing automation and building a manual search system.
How we transformed a women’s fashion brand from unstable ROAS to 15.74× peak — while cutting ad budget by 55%. A 4-year Meta Ads partnership.
How we built a Meta Ads system from zero for an AI fintech startup — CTR 3%→13%, CPA $21→$8, conversion rate up 4×.
How we achieved 797% ROAS for a board games store through campaign restructuring and Server-Side Tracking — up from 250% under a previous agency.
How we scaled a SaaS Black Friday campaign to 430% ROAS with $177K budget on Google Ads — Search and Performance Max.
How we scaled a sports merchandise Shopify store from $5K to $25K/mo ad spend while maintaining 4x+ ROAS through feed segmentation and multi-channel expansion.