Return Windows Are Distorting Your Google Ads Bidding

The Return Window Gap

Every ecommerce business operates with a return window - typically 14 to 60 days. During this window, any sale Google Ads claims credit for might be reversed. But Google's attribution model records the conversion at the point of purchase, not at the point where the return window closes.

This creates a structural gap between what Google reports and what your bank account shows. The gap isn't small. UK fashion retailers see return rates of 25-40%. Footwear runs 20-35%. Even homewares - traditionally a low-return category - sits at 8-15% for online orders.

The problem compounds because return rates aren't uniform across products. Your best-selling dress might have a 35% return rate while your best-selling candle has a 3% return rate. If both show the same ROAS in Google Ads, the candle is dramatically more profitable - but Smart Bidding treats them identically.

This isn't a minor reporting inconvenience. It's a systematic bias that directs budget toward products that look profitable on paper but erode margin when returns are processed. The return rate should be a core input to your bidding strategy, not an afterthought reconciled quarterly.

How Returns Distort ROAS

Consider two products in the same Google Ads account:

• • Product A: £80 AOV, 50% gross margin, 5% return rate. Net revenue per sale: £76. Gross profit: £38. Post-return gross profit: £36.10.
• • Product B: £120 AOV, 45% gross margin, 30% return rate. Net revenue per sale: £84. Gross profit: £37.80. Post-return gross profit: £26.46.

Google Ads shows Product B generating higher revenue per conversion (£120 vs £80) and similar gross margin. Smart Bidding allocates more budget to Product B because it appears to generate more conversion value. But after returns, Product A delivers 36% more gross profit per sale.

Over a month, if each product receives £5,000 in ad spend and generates 100 conversions, Product B's reported ROAS is 2.4x while Product A's is 1.6x. The agency celebrates Product B's "stronger performance." But Product A generated £3,610 in actual gross profit vs Product B's £2,646. Product A is 36% more profitable despite "worse" ROAS.

This distortion is invisible in standard Google Ads reporting. It requires joining conversion data with return data at the product level - something most agencies never do because they don't have access to your returns system.

The Timing Mismatch

Smart Bidding optimises in near real-time. It makes bid decisions based on conversion data from the last 7-30 days. But your return window might be 30-60 days. This means Smart Bidding is always making decisions on incomplete data - it's seeing the sales but not yet seeing the returns.

The timing mismatch is worst after promotional periods. Black Friday generates a spike in conversions, making Smart Bidding bullish on the products and audiences that converted. But promotional buyers return at higher rates - often 40-50% higher than organic buyers. By the time returns materialise in January, Smart Bidding has already allocated more budget based on the inflated November data.

You can partially address this by feeding return data back into Google Ads via the conversion adjustment API. But most brands don't do this. And even those that do face a structural lag - you can only upload adjustments after returns are processed, by which point Smart Bidding has already made thousands of bid decisions based on the original data.

The practical solution is proactive: build return rate assumptions into your targets before Smart Bidding makes its decisions, rather than correcting retroactively.

Category-Level Return Rates

Return rates vary dramatically by category, and this variance should shape how you structure campaigns and set targets:

• • Fashion/apparel: 25-40% return rate. Size, fit, and colour are primary drivers. Multi-size ordering inflates returns further. This category needs the most aggressive return-adjusted targets.
• • Footwear: 20-35%. Fit is the primary driver. Width options and half-sizes can reduce returns but increase variant complexity in your feed.
• • Electronics/accessories: 5-12%. Lower returns but higher return processing costs. "Not as described" returns often come from poor product imagery.
• • Home & living: 8-18%. Size perception issues (furniture photos that mislead on scale) and colour accuracy drive returns. Bulky items have higher return logistics costs.
• • Beauty/skincare: 3-8%. Low return rates but "tried and didn't like" returns are non-resaleable, meaning 100% margin loss on returned items.

If you run a multi-category retailer with a single blended ROAS target, you're implicitly cross-subsidising high-return categories with low-return categories. Your homewares margin is funding your fashion returns. Separate campaigns with category-specific, return-adjusted targets fix this.

Building Return-Adjusted Targets

The calculation is straightforward but rarely done:

• • Step 1: Calculate your post-return revenue per sale. If AOV is £100 and return rate is 25%, post-return revenue is £75.
• • Step 2: Calculate post-return gross profit. If gross margin is 50%, pre-return profit is £50. Post-return profit is £37.50 (£75 × 50%).
• • Step 3: Factor in return processing costs. Shipping, inspection, restocking, and customer service costs per return might be £8. With 25 returns per 100 sales, that's £200 in return costs, reducing profit to £35.50 per sale.
• • Step 4: Set your ROAS/POAS target based on post-return economics. If you need £30 profit per sale to cover ad spend plus overhead, your minimum POAS is £35.50/CPA. Your ROAS target should be £100/CPA (not £75/CPA) because Google reports the pre-return figure.

The adjusted target formula: Adjusted ROAS target = Base ROAS target ÷ (1 - return rate). If your base target is 4x and the return rate is 25%, your adjusted target is 5.33x.

This single adjustment prevents the systematic overbidding that occurs when return rates are ignored. Apply it at the campaign or product group level, using historical return rates from your ecommerce platform - not from Google Ads, which doesn't have this data.

When to Bid Harder Despite Returns

Return-adjusted bidding doesn't always mean bidding less. Some high-return products are worth aggressive bidding because:

• • Exchange rate is high: If 60% of "returns" convert to exchanges (different size/colour), you retain the revenue and the customer relationship. The net return rate is much lower than the gross return rate.
• • Returners still buy: Customers who return one item but keep two others from the same order are still profitable. Your return rate at the item level doesn't capture order-level profitability.
• • LTV justifies first-order losses: Fashion brands with strong repeat purchase rates can tolerate high first-order return rates because lifetime value makes up the difference. The return is a customer acquisition cost, not a pure loss.
• • Returnable products drive discovery: Some products (like multi-packs or trial sets) have high return rates but introduce customers to your brand. The returned items are the cost of customer acquisition.

The distinction requires nuance. Blanket return-rate adjustments are better than no adjustments, but the best approach segments by return type (refund vs exchange), return reason (fit vs quality vs preference), and customer behaviour (first-time vs repeat buyer). This level of granularity separates competent account management from genuine commercial strategy.

Next Steps