Why 800% ROAS Can Lead to Losses and How to Implement Profit-Based Bidding

Every digital agency prides itself on high ROAS. If a company invests 5,000 euros in advertising and generates 40,000 euros in revenue, the ROAS is 800% (or 8x). The report looks excellent, management is satisfied, and the marketing specialist requests a budget increase. However, at the end of the month, the company's CFO announces: "We don't have free funds to pay the bills, and the company operated at a loss this month."

This is the classic e-commerce and digital business paradox created by blind reliance on superficial revenue metrics. Google Ads and Meta algorithms do not know your cost of goods sold (COGS), warehouse expenses, payment card fees, or return rates. If a campaign is optimized for simple revenue, the algorithm begins to mass-sell products that are easy to sell but have negligible profit margins. The solution is strategic profit-based Google Ads management and a transition to the POAS model, supported by integrated data analytics and attribution.

What is POAS and How It Differs from Traditional ROAS

To understand the difference, let's look at the mathematical structure of both formulas:

• ROAS (Return on Ad Spend) = Revenue ÷ Ad Spend: Shows only how many euros of revenue are generated for every euro spent on advertising. It completely ignores cost of goods and profitability.
• POAS (Profit on Ad Spend) = Gross Profit ÷ Ad Spend: Shows how many euros of actual gross profit (after deducting COGS and shipping) are returned for every euro invested in advertising.

The POAS threshold is crystal clear: if POAS is above 1.0, the campaign is generating real net profit. If POAS is below 1.0, every subsequent order directly burns the company's money, regardless of whether the ROAS is 500% or 1500%.

Mārtiņš Grūbe Digital Marketing & Google Ads Strategist

"Revenue is ego, profit is reality. If your best campaign sells products with a 7% margin and generates 800% ROAS, you are losing money on every purchase due to logistics and advertising costs. Only by teaching the Google algorithm to receive actual gross profit figures does marketing become a stable profit machine."

Real-World Example: How 800% ROAS Generated 2,400 Euro Losses

Let's look at a specific e-store situation with two different product groups and an identical advertising budget of 2,000 EUR each:

• Campaign A (Low-margin electronics): Sells popular brand smartphones with an 8% gross margin. An ad budget of 2,000 EUR generated 16,000 EUR in revenue. ROAS = 800% (8x). However, the gross profit was only 1,280 EUR (8% of 16,000). Subtracting ad costs (1,280 - 2,000), the net result: -720 EUR loss!
• Campaign B (High-margin accessories and service): Sells specialized accessories with a 55% gross margin. An ad budget of 2,000 EUR generated 7,000 EUR in revenue. ROAS = 350% (3.5x). Gross profit was 3,850 EUR (55% of 7,000). Subtracting ad costs (3,850 - 2,000), the net result: +1,850 EUR real profit!

A campaign manager evaluating only ROAS would have reduced the budget for the 'weak' Campaign B and given all the money to the 'excellent' Campaign A, thereby driving the company deeper into the red. Only a POAS audit allowed them to see this catastrophe.

Contribution Margin Levels (CM1, CM2, and CM3)

To correctly implement POAS, three levels of Contribution Margin must be distinguished in the company's financial model:

• CM1 (Contribution Margin 1): Revenue minus direct Cost of Goods Sold (COGS). This is the base level usually used in simple e-stores.
• CM2 (Contribution Margin 2): CM1 minus shipping expenses, packaging costs, and payment card commissions (Stripe, banklink). This metric provides a much truer picture of each basket's profitability.
• CM3 (Contribution Margin 3): CM2 minus return processing costs and customer service overhead.

By passing the CM2 value to the advertising system via Server-Side CAPI, the Google Ads machine learning algorithm is trained to sell products that leave the company with the most free cash flow after covering all variable costs.

Product Feed Segmentation with Margin Custom Labels in Google Shopping

In addition to changing bidding strategies, the best practice is to segment the Product Feed using Custom Labels:

• Custom Label 0 - High Margin (> 40%): A lower Target ROAS is allowed here (e.g., 250-300%), allowing the campaign to aggressively capture the market because every transaction brings high net profit.
• Custom Label 0 - Medium Margin (20-40%): Balanced Target ROAS (e.g., 450-550%).
• Custom Label 0 - Low Margin (< 15%): Strict Target ROAS (e.g., 800-1100%), preventing the algorithm from selling these products at a loss.

This segmentation prevents low-cost products from dominating reports and guarantees that every invested advertising euro works for the company's true profit.

Step-by-Step Guide to Implementing POAS Bidding in Google Ads

Transitioning to profit-based bidding involves four technical steps:

1. COGS Integration in the Database: A specific purchase price is linked to each product in the e-store system or ERP.
2. Gross Profit Calculation at Checkout: When a customer places an order, the server calculates the actual profit: Gross Profit = Sale Price - COGS - Shipping Expenses.
3. Passing Profit as Conversion Value via Server-Side CAPI: Instead of the standard basket total, the calculated profit figure is passed to the advertising system in the value parameter. This is handled by a secure Server-Side CAPI connection.
4. Target ROAS Calibration to POAS Target: The Google Ads bidding goal is transformed: instead of asking for 800% of revenue, a goal is set to achieve 140-160% of the profit value (Target POAS 1.4-1.6).

Shifting Management Metrics in the Boardroom and Incentive Systems

The shift from ROAS to POAS also requires a mindset change at the executive level. Traditionally, marketing teams and agencies are paid bonuses for revenue growth. As long as bonuses are tied to revenue, it is beneficial for marketing to spend money on low-margin promotional items because they grow total income more easily.

By implementing POAS as the primary KPI, the goals of marketing teams and agencies become 100% synchronized with the owner's interests: marketing is rewarded for real money earned, not for throughput revenue. This automatically eliminates campaigns that burn budget without profit.

Results Comparison: Before and After POAS Implementation

See how the company's metrics changed over three months after restructuring the campaigns:

How to Automate Profit Calculation in Shopify and WooCommerce

For the POAS model to work without manual labor, an automated script is connected to the e-commerce platform that reads purchase prices from the database for every order. In Shopify, the Cost of Goods Sold (COGS) field is used, while in WooCommerce, a simple custom field like _cost_price is utilized.

At the moment of purchase, a post-processing Webhook calculates the real profit and automatically sends it to Google Ads and Meta CAPI. Accounting and marketing data finally become a unified whole, and campaign management gains full financial control over every invested euro.

Category Budget Reallocation: The POAS Decision Matrix

Every month, the campaign manager compiles POAS metrics for product categories. Categories with a POAS exceeding 1.8 receive a 20% budget increase. Categories with a POAS between 1.0 and 1.4 are optimized by seeking better suppliers or reducing CPC. Meanwhile, categories with a POAS below 1.0 are immediately paused in advertising, protecting the company from hidden losses.

Price Elasticity Testing and Ad Cost Impact on Profitability

A common phenomenon in e-commerce: a small price increase of 5% can double a product's net profit, even if sales volume decreases by 2%. POAS tracking allows for safe price elasticity experiments, monitoring in real-time how price changes affect campaign profitability and total profit mass over the month.

Common Discrepancies Between Accounting and Marketing Data

When switching to POAS, discrepancies often emerge between conversions registered in Google Ads and actual bank income. The main reasons are cancelled orders, returned goods, and unpaid invoices. By implementing a 14-day delayed data correction protocol, these negatives are automatically deducted from campaign value, ensuring the algorithm does not rely on illusory data.

Inventory-Driven Bidding: Dynamic Bidding Based on Stock Levels

POAS efficiency is also significantly affected by warehouse stock levels. If a company has only 3 units left of a high-profit item, there is no point for the Smart Bidding algorithm to continue aggressive spending on that item, as it will sell out organically. By integrating warehouse stock data with the ad feed, campaigns automatically lower bids for items with low stock and redirect budget to items where faster turnover is needed.

This integration combines marketing with supply chain management, ensuring the company's working capital is used with maximum return without unnecessary capital freezing in the warehouse.

Frequently Asked Questions About POAS and Profit Bidding