The instinct when an ad account underperforms is to fix the creative, adjust the bid, or test a new audience. The actual problem, most of the time, is the catalog. A meaningful slice of SKUs is running ads that had no realistic path to conversion: broken sizes, low-value items, products that attract clicks but don't close. For brands managing $50K-$150K/month across 1,000+ SKUs, the answer isn't more manual review. It's automated guardrails that stop waste before it compounds, and a discipline for redirecting what's saved.
At AdYogi, we manage over $150M in ad spend across 350+ brands. This guide covers how to configure Stop Loss, Smart Products Exclusion, Product Performance Tracking, and ABO through BigAtom to cut that waste systematically.
Across AdYogi's managed portfolio, the average large-catalog brand is losing 20-25% of monthly budget to SKUs that never had a realistic chance to convert. Meta and Google's algorithms distribute impressions broadly. Without a guardrail telling the platform which products are worth backing, that distribution includes your broken inventory, your low-value filler, and products that attract clicks because the image looks good but fail at checkout for reasons no bid adjustment will fix.
The specific culprits are consistent across accounts:
Manually auditing thousands of SKUs daily to catch and pause these is operationally impossible for most teams. The brands that solve it build automated systems. The brands that don't keep hiring for the problem.
To automate your optimization, you need a clear performance threshold first. We use ACOS (Advertising Cost of Sales) as the primary efficiency metric.
AdYogi's optimization engine is performance-based, not margin-based. The platform does not ingest Cost of Goods Sold (COGS) data or directly measure contribution margins. It uses ACOS and conversion rates as direct proxies for performance.
At catalog scale, ACOS is the right signal. You're running automated rules across thousands of SKUs simultaneously. The signal has to work without manual interpretation, and ACOS does.
Starting thresholds vary by category and lifecycle stage. Calibrate from here:
| Product Category | Typical Gross Margin | Target ACOS Threshold (Starting Point) | Evaluation Window |
|---|---|---|---|
| Core Apparel (High Margin) | 60% - 70% | 30% - 35% | 7 Days |
| Footwear / Accessories | 50% - 60% | 25% - 30% | 7 Days |
| Fine Jewelry / Premium | 40% - 50% | 20% - 25% | 14 Days |
| Clearance / End-of-Season | 30% - 40% | 15% - 20% | 3 Days |
These thresholds are illustrative starting points. Every brand must calibrate these numbers based on their specific business model, average order value (AOV), and customer lifetime value (LTV).
Stop Loss is an automated rule engine within BigAtom. At the campaign, adset, ad, or product level, it watches performance against your defined thresholds and pauses anything that breaches them before the waste accumulates.
Strategic note: Stop Loss is a guardrail, not a strategy engine. It prevents waste by cutting off underperforming assets. It does not replace strategic product decisions, creative testing, or audience positioning.
The mechanism is straightforward: underperforming products don't get the weekend to keep burning spend. In a client-approved case study, Aza Fashion used Stop Loss rules to automatically pause non-performing assets, saving up to 25% of their monthly ad spend.
For Libas (women's ethnic fashion, 5,000+ SKU catalog), AdYogi's stop-loss system identified optimal spend thresholds per SKU within each product segment. When a SKU's performance began to decay, promotion was automatically halted and budget was reallocated to the next-best performing products in the segment. The savings were tracked and quantified, contributing to Libas growing from Rs 60 crore to Rs 300 crore in revenue over three years with AdYogi.
The system runs across every SKU simultaneously. A weekly review catches last week's waste. Stop Loss catches it before the weekend burns.
Stop Loss is effective, but it's your second line of defense. AdYogi's first line of defense is Smart Products Exclusion, which cleans your catalog feed before Stop Loss ever engages.
This module, part of the BigAtom platform, automatically filters out low-potential or broken products from your active ad feeds based on inventory and feed health. It prevents Meta and Google from spending money on products that are fundamentally unpurchasable or visually unappealing.
No. Smart Products Exclusion only removes products from your active advertising feeds (Meta and Google catalogs). It does not delete products from your Shopify, Magento, or WooCommerce backend, nor does it remove them from your website's organic navigation. If a product's inventory is replenished, AdYogi's hourly catalog sync will automatically restore the product to the active ad feed, typically within the hour.
Once the guardrails are live, you need visibility into what they're catching. AdYogi's Product Performance Tracking module compares ad spend per product against two key performance indicators, conversion rate and ACOS, giving growth leads a weekly view of exactly where budget is working and where it isn't.
| Product Name | Ad Spend | Conversion Rate | Product ACOS | Action |
|---|---|---|---|---|
| Floral Maxi Dress | $4,200 | 3.2% | 22% | Scale |
| Linen Summer Shirt | $1,800 | 1.1% | 48% | Optimize |
| Silk Scarf | $950 | 0.4% | 85% | Pause |
By reviewing this dashboard weekly, growth leads can identify:
Reminder: Product Performance Tracking measures ACOS and conversion rate. It does not measure gross or contribution margin directly.
Once Stop Loss and Smart Products Exclusion have cut the waste, that freed budget needs to go somewhere productive. AdYogi's Automatic Budget Optimizer (ABO) takes the reclaimed spend and puts it to work.
ABO checks your active campaigns on a scheduled basis and shifts budget toward campaigns, adsets, and products that are meeting or exceeding your target ACOS. The reallocation logic runs on an automated schedule, not instant triggers, which keeps the system stable across normal intraday fluctuation.
The budget recovery this enables is meaningful. For Sureena Chowdhri (luxury designer apparel, AOV Rs 18,000-22,000), AdYogi identified that 15-20% of total ad budget was flowing to low-intent geographies and underperforming placements. By cutting those placements and redirecting that spend to higher-converting product sets via BigAtom's Smart Product Segments (BigAtom's SKU-tier grouping), Meta ROAS was maintained even as total ad spend increased by 50%. That reclaimed budget funded growth directly. Sureena Chowdhri scaled monthly online revenue 6X in six months, from approximately Rs 50 lakh to Rs 3 crore, with AdYogi.
AdYogi optimizes performance by selecting which SKUs to back (i.e., which products are included in active ad sets and catalogs). Actual bid management and delivery optimization are delegated to Meta and Google's native machine-learning algorithms. This hybrid approach combines AdYogi's catalog control with the platforms' bidding engines. AdYogi selects which products get backed. Meta and Google handle the bidding. Neither system is overriding the other.
To eliminate ad spend waste and scale your D2C fashion brand profitably, implement AdYogi's four-part BigAtom workflow in this order:
Most in-house teams don't have the bandwidth to maintain this granularity week after week. The gap between knowing the framework and running it consistently is where profit gets left on the table.
Deploy Automated Guardrails