Product vs Channel Profitability

ROAS measures attributed revenue per ad dollar. A product-channel combination also carries its own average order value, product mix, fulfillment, fees, returns, and other variable costs. Rank comparable scenarios by the contribution left after those costs and ad spend.

Scenario revenue = ad spend x attributed ROAS

Normalized orders = scenario revenue / average order value

Contribution profit after ads = scenario revenue x contribution margin - ad spend

Equal spend controls only the amount invested in each scenario. It does not equalize impressions, clicks, conversion rate, order availability, customer quality, attribution rules, or incrementality.

Input	Possible evidence	Required boundary
Attributed revenue and ROAS	Channel report using a documented window and value field	Attribution is reporting credit, not causal proof
AOV and item mix	GA4 ecommerce events and selected Data API metrics/dimensions	Use the same revenue and refund basis
Product unit cost	Shopify InventoryItem unitCost when populated and current	Unit cost is not complete contribution cost
Other variable costs	Fulfillment, payment, marketplace, returns, duties, packaging, and service ledgers	Include costs caused by the order or channel
Ad spend	Matched channel, product set, dates, timezone, and currency	Do not mix different spend populations

Shopify unitCost can support merchandise-cost mapping, but it does not include the complete product-channel cost boundary. Add fulfillment, payment and marketplace fees, return loss, duties, packaging, incentives, and other variable costs before treating the result as contribution margin.

1. Freeze one period, timezone, currency, advertising attribution setting, and refund-maturity date.
2. Build one compatible Shopify data exploration at product, variant, and Shopify sales-channel grain. Preserve Product ID and Product variant ID when compatible, product and variant labels, SKU, gross sales, discounts, Sales reversals, net sales, net quantity, matched Order ID or same-grain Orders, COGS, and net sales with and without cost recorded.
3. Export the matched advertising rows separately with ad product or product-set key, attributed revenue, ad spend, dates, timezone, currency, and attribution setting. Shopify Sales channel identifies where an order was placed; it is not an advertising attribution channel.
4. Create an explicit crosswalk for the extract period. Prefer Shopify Product variant ID mapped to the ad item key; use SKU only after proving it is populated, unique, and stable. Titles are display labels, not fuzzy join keys, and a replacement variant can have a new ID.
5. Add fulfillment, fees, reverse-logistics loss not already represented in net sales or COGS, duties, packaging, and other order-driven costs from their source ledgers.

Compatibility boundary: Shopify disables incompatible dimensions and metrics while a report is edited. Exact columns depend on the compatible exploration you select; do not assume every field appears in one default export. If split extracts lack a shared lossless key such as Order ID, Sale ID, or Product variant ID, stop instead of inferring a product-by-channel result from separate totals.

Cost-coverage gate: A missing product cost can make Shopify COGS display as $0. Check net sales with cost recorded against net sales without cost recorded and stop the scenario when material revenue lacks cost coverage; never convert missing cost into a verified zero.

Aggregate	Ratio-of-sums calculation	Scenario parameter
Period contract	Period + currency + timezone + advertising attribution setting	period
Scenario label	Non-sensitive product-channel alias	s#n
Modeled scenario AOV	Sum mature net sales / count unique matched orders	s#a
Contribution margin	(Sum net sales - COGS - fulfillment - fees - uncovered reverse logistics - other variable costs) / sum net sales	s#m
Attributed ROAS	Sum attributed revenue / sum matched ad spend	s#r
Ad spend	Sum matched ad spend	s#s

Never average the row ratios. Add their numerators and denominators first. This net-sales-per-order value is ROAS Break's modeled scenario AOV, not Shopify's official Average order value metric. Net quantity is not an order count; unless one item is proven to equal one order, it cannot supply the denominator. If orders contain multiple selected products, deduplicate Order ID within each scenario before counting.

Sales reversals are not physical returns. Shopify uses Sales reversals for the monetary value reversed through returns, cancellations, and order edits; it is the successor to the legacy Returns metric. Keep physical return handling and reverse-logistics loss in the external cost ledger.

Privacy boundary: Keep raw exports in your controlled workspace. Put only aggregate values and non-sensitive aliases in the shareable Scenario Planner URL; never include SKU, product or variant ID, order ID, customer data, or source filenames. This workflow does not upload a file.

Copy Shopify mapping worksheet

This fictional two-row extract shows why every ratio must be rebuilt from totals. Contribution before ads uses the same net-sales and cost boundary in both rows.

Row	Net sales	Unique orders	Contribution before ads	Attributed revenue	Ad spend
Week 1	$12,000	200	$3,000	$12,000	$4,000
Week 2	$28,000	300	$13,000	$28,000	$6,000
Combined inputs	$40,000	500	$16,000	$40,000	$10,000

Modeled scenario AOV = ($12,000 + $28,000) / (200 + 300) = $80.00

Contribution margin = ($3,000 + $13,000) / ($12,000 + $28,000) = 40.00%

Attributed ROAS = ($12,000 + $28,000) / ($4,000 + $6,000) = 4.00x

A simple average would instead report $76.67 AOV, 35.71% margin, and 3.83x ROAS. Those values are wrong because the rows have different denominators. The ratio-of-sums result maps to s2a=80, s2m=40, s2r=4, and s2s=10000.

Give three fictional product-channel combinations the same $10,000 ad spend. The margin is contribution before advertising, not gross margin. Each scenario assumes its entered AOV, contribution margin, and attributed ROAS hold at that spend level.

Scenario	AOV	Contribution margin	ROAS	Revenue	Normalized orders	Contribution profit
High ROAS low margin	$50	20%	5.00x	$50,000	1,000	$50,000 x 20% - $10,000 = $0
Balanced mix	$80	40%	4.00x	$40,000	500	$40,000 x 40% - $10,000 = $6,000
Lower ROAS high margin	$100	55%	3.00x	$30,000	300	$30,000 x 55% - $10,000 = $6,500

The 5.00x scenario merely breaks even after advertising because its 20% contribution margin supplies exactly $10,000 for the $10,000 spend. The 3.00x scenario wins because 55% of $30,000 creates $16,500 of contribution before ads and leaves $6,500 afterward.

Orders are normalization only: The planner divides revenue by AOV. The resulting 1,000, 500, and 300 are not imported order counts and should not replace observed transactions when those are available.

Compare the three product-channel scenarios

If Lower ROAS high margin falls from 3.00x to 2.80x at the same $10,000 spend, revenue becomes $28,000, normalized orders become 280, and contribution profit becomes $28,000 x 55% - $10,000 = $5,400. Balanced mix remains at $6,000 and becomes the winner.

Stress test	Revenue	Normalized orders	Contribution profit	Rank
Balanced mix at 4.00x	$40,000	500	$6,000	Winner
Lower ROAS high margin at 2.80x	$28,000	280	$5,400	Second
High ROAS low margin at 5.00x	$50,000	1,000	$0	Third

Do not scale this ranking linearly. More spend can change auction price, traffic quality, conversion rate, AOV, stock availability, discounting, fulfillment cost, return rate, and product mix. Recalculate marginal contribution at realistic spend steps instead of multiplying the current winner.

Attributed revenue is not proof of incremental revenue. Use experiments or another credible causal design when the decision depends on lift. If multiple channels credit the same order, reconcile those claims separately rather than adding scenario revenue.

1. Define each product-channel pair and freeze report dates, attribution settings, timezone, currency, and revenue field.
2. Export observed product and transaction evidence from GA4 and the commerce platform.
3. Map Shopify unit cost where useful, then complete the variable cost boundary outside that field.
4. Calculate scenario-specific contribution margin and document whether refunds are mature.
5. Compare at equal spend, then stress-test ROAS, AOV, margin, and spend independently.
6. Choose the action from total and marginal contribution, inventory capacity, and uncertainty, not ratio rank alone.

Boundary: The POAS guide defines profit-on-ad-spend metrics. The attribution-versus-MER guide reconciles channel credit with the store. The Shopify guide maps report fields. This page uses already-defined inputs to rank product-channel combinations on total contribution profit.

Google documents ecommerce collection and Data API reporting fields. Shopify documents analytics dimensions and metrics, compatibility controls, sales and profit reports, report exports, SKU behavior, InventoryItem, and unitCost. Amazon Ads supplies the ROAS terminology used in the scenario. ROAS Break supplies the manual crosswalk, ratio-of-sums method, privacy boundary, fictional inputs, equal-spend control, and profit ranking.

• Google Analytics: Measure ecommerceOfficial transaction, item, purchase, and refund collection context. Checked August 21, 2026.
• Google Analytics Data API schemaOfficial reporting dimensions and metrics for selecting a consistent extract. Checked August 21, 2026.
• Shopify analytics fieldsOfficial definitions for product, variant, sales-channel, sales-reversal, order, and cost-coverage fields. Checked August 21, 2026.
• ShopifyQL sales schemaOfficial sales dataset dimensions, metrics, and net-sales relationships. Checked August 21, 2026.
• Shopify: Filter and edit reportsOfficial metric and dimension compatibility behavior for data explorations. Checked August 21, 2026.
• Shopify sales reportsOfficial product, variant, sales-channel, net-sales, and sales-reversal reporting context. Checked August 21, 2026.
• Shopify profit reportsOfficial product-cost, net-sales, COGS, gross-profit, and cost-coverage context. Checked August 21, 2026.
• Shopify: Export reportsOfficial report export workflow and CSV context. Checked August 21, 2026.
• Shopify product CSV guidanceOfficial product and variant ID lifecycle warning when variant options are changed by CSV. Checked August 21, 2026.
• Shopify SKU guidanceOfficial SKU purpose, uniqueness, and maintenance guidance. Checked August 21, 2026.
• Shopify Admin GraphQL: InventoryItemOfficial InventoryItem fields, including unitCost. Checked August 21, 2026.
• Amazon Ads ROAS guideOfficial ROAS definition and interpretation. Checked August 21, 2026.