Industry Guide Consultancy
Revenue growth can hide unprofitable orders. Build contribution margin per order after returns, support and acquisition, segment it by channel and product, then test the levers.
Many online stores know their revenue, gross margin and return on ad spend, yet still cannot say whether a typical order makes money. The reason is that profitability is spread across teams. Marketing sees acquisition costs. Operations sees fulfillment and shipping. Customer service sees support costs. Finance sees the total at month end, after it is too late to change anything.
E-commerce profitability planning brings those costs together at the level where decisions are made: the order, the customer, the channel and the product. This guide shows how to calculate contribution margin after returns and support, segment it, model promotions and repeat purchases, and choose improvement scenarios.
Returns are not a rounding error. The National Retail Federation and Happy Returns' 2025 Retail Returns Landscape report, based on surveys of US consumers and of e-commerce professionals at merchants with more than $500 million in revenue, projected that retailers would see 15.8% of annual sales returned in 2025, with an estimated 19.3% of online sales returned. Retailers in the survey expected 17% of holiday sales to come back, and estimated that 9% of all returns were fraudulent.
Each return can carry reverse shipping, handling and inspection costs, items that cannot be resold at full price, payment fees that are not refunded and customer service contacts. A product with a healthy gross margin and a high return rate can lose money on every order.
Contribution margin is what remains from revenue after all variable costs of fulfilling an order. A practical e-commerce version has three layers:
The following figures are invented for illustration. Replace them with your own data.
A store sells apparel and accessories with an average order value of $80 before returns. Product cost is 40% of retail price. The return rate is 18%, and 80% of returned items can be resold.
| Line (per order) | Calculation | Amount |
|---|---|---|
| Gross order value | — | $80.00 |
| Net revenue after returns | $80 × (1 − 18%) | $65.60 |
| Cost of goods sold | Kept items $26.24 plus unsellable returns $1.15 | −$27.39 |
| Payment fees | 3% of $80, assumed not refunded on returns | −$2.40 |
| Fulfillment and packaging | Assumed | −$4.50 |
| Outbound shipping (net of shipping charged) | Assumed | −$7.00 |
| Return handling and reverse shipping | 18% × $11 per return | −$1.98 |
| Customer support | 0.15 contacts per order × $5 per contact | −$0.75 |
| Contribution before marketing | $21.58 |
Now add marketing. Suppose a new customer costs $35 to acquire, and retention marketing for repeat orders costs about $3 per order.
If 40% of new customers return within twelve months and those who do place an average of 1.6 repeat orders, the expected repeat orders per new customer is 0.64. Twelve-month contribution per new customer is then about −$1.53. The store is growing revenue while losing a little on each new customer in the first year.
This is a common pattern, and it is not necessarily a problem if customers keep buying beyond twelve months. But it should be a decision, not a surprise.
Most stores already hold the data needed; it simply sits in different systems:
| Cost line | Typical source |
|---|---|
| Revenue, discounts and refunds | Commerce platform and payment processor reports |
| Product cost | Inventory or accounting system, ideally landed cost including inbound freight and duties |
| Payment fees | Payment processor statements |
| Fulfillment and packaging | Third-party logistics invoices or internal labor and materials costs |
| Shipping and return shipping | Carrier invoices, matched to orders where possible |
| Support costs | Helpdesk contact volumes and support team costs, or outsourcing invoices |
| Marketing | Ad platforms, affiliate and agency invoices, matched to new and returning customers |
Start with monthly averages if order-level matching is not yet possible. An approximate model that the whole team trusts is more useful than a precise one that never gets finished.
A few decisions change the answer, so agree them with finance up front:
Averages hide where profit is made and lost. Recalculate the model for segments that matter:
A simple matrix of contribution margin 3 by channel and category usually reveals a few segments that fund the business and a few that drain it.
Discounts reduce revenue on every unit while most costs stay the same. In the example above, a 20% discount reduces the price from $80 to $64. After returns, fees and the same fulfillment, shipping and support costs, contribution before marketing falls from $21.58 to about $8.94 per order, a drop of nearly 60%. To earn the same total contribution, the promotion would need to generate roughly 2.4 times as many orders, or bring in customers who are unusually likely to buy again.
Before running a promotion, model three things: the incremental orders it must create, the share of discounted orders that would have happened anyway, and the repeat purchase behavior of customers acquired through discounts. Our Black Friday email and SMS checklist applies this to peak-season campaigns.
Repeat purchase is usually the largest lever on e-commerce profitability because it spreads acquisition cost across more orders. Track repeat rate and orders per customer by cohort, and test what moves them: post-purchase communication, replenishment reminders, loyalty benefits and product recommendations.
A cohort table makes this visible. List customers by the month of their first order, then show cumulative contribution per customer at three, six, twelve and twenty-four months. Cohorts that reach break-even quickly show which channels, products and offers attract valuable customers. Cohorts that never reach break-even show where acquisition spending should be reduced, however strong their first-order revenue looked at the time.
With the model built, compare levers by their effect on twelve-month contribution per new customer. Using the same hypothetical store:
| Scenario | Change | Contribution before marketing | 12-month contribution per new customer |
|---|---|---|---|
| Baseline | — | $21.58 | −$1.53 |
| Reduce returns | Return rate from 18% to 14% through better size guidance and product information | $24.19 | +$2.76 |
| Raise order value | Average order value from $80 to $90 through bundles, other costs unchanged | $26.05 | +$5.81 |
| Lower acquisition cost | From $35 to $28 through channel mix | $21.58 | +$5.47 |
| Heavy discounting | 20% off all orders, no change in behavior | $8.94 | −$22.26 |
Each scenario holds other assumptions constant, which real life will not. The value of the table is in ranking levers and starting the right conversations: in this example, returns reduction and order value improvements are worth as much as a substantial cut in acquisition cost.
Returns: clearer size guidance, better product photos and descriptions, fit feedback from reviews, policies that discourage serial returning, exchanges instead of refunds, and fraud checks informed by data on unusual return patterns.
Order value: bundles, thresholds for free shipping set with margin in mind, and relevant add-ons at checkout.
Acquisition: shift budget toward channels and campaigns with the best contribution margin 3, not the best return on ad spend; exclude existing customers from prospecting campaigns.
Support: reduce avoidable contacts with proactive order updates and clear returns information.
Fulfillment and shipping: packaging that fits, carrier negotiation and zone-based shipping strategies.
Once the model is in place, use it in annual and quarterly planning. Set targets for contribution per order and per new customer alongside revenue, give each lever an owner and a target, and review progress monthly. When leadership debates a new channel, product line or promotion, ask for its expected effect on contribution margin 3, not only on revenue. Over time, that habit changes which ideas get funded.
There is no single good number, because margins vary by category, price point and business model. A more useful test is whether each order covers product cost, fulfillment, payment fees, returns and support, and whether each new customer's twelve-month contribution exceeds the cost of acquiring them.
Start with net revenue after discounts and returns and subtract product cost (contribution margin 1), then subtract variable fulfillment, payment, returns handling and support costs (contribution margin 2), then subtract marketing costs (contribution margin 3). Calculate it per order, by channel and by product category.
Acquisition cost lands on the first order. In this guide's example, an order contributes $21.58 before marketing, so a $35 acquisition cost makes the first order lose $13.42. Profit then depends on how many customers return and how often they buy again.
Build the per-order model for your top three channels and top five categories using last quarter's data. Most teams find at least one segment where they are paying to lose money, and one where they are under-investing.
Growix's Financial Planning & Business Modelling team builds the forecasts and models that turn plans into numbers you can manage against, and our Growth Strategy team helps decide where to grow and what to stop. If support costs are a major line in your model, see our guide to outsourced customer support pricing.
Most SaaS plans fail because the revenue target, pipeline math, retention and delivery capacity were set separately. A worked model that connects them, plus a quarterly review.
AI shopping tools can only recommend what they can read and trust. Make product facts complete, consistent and crawlable across pages, feeds and markup, then measure sales, not mentions.
The useful e-commerce agent is not the most autonomous one. Match each workflow to an autonomy level, connect the right systems, and measure correct actions, not just deflection.
