Your Bestseller Might Be the SKU Weakening the Business

The product with the most revenue is not automatically the product creating the most value.

It may carry the deepest discount. It may be expensive to fulfil. It may attract customers who rarely return. Its size or fit may create an unusually high return rate. Paid campaigns may keep it at the top of the sales report while consuming most of the contribution it appears to generate.

None of that makes it a bad product. It means “bestseller” is not a commercial conclusion.

Before putting more media budget or inventory behind a SKU, a growing Shopify brand should be able to answer:

What remained after this product absorbed the costs required to sell, fulfil and keep the order?

Revenue hides the route to profit

Shopify is a useful commerce source of truth. It records the order, line item, discount, refund and product. But the decision to promote a product also depends on costs held elsewhere:

• current product cost or landed cost;
• pick-and-pack and outbound shipping;
• payment or marketplace fees;
• return handling and write-offs;
• paid-media spend;
• the customer’s subsequent orders, if repeat behaviour matters to the decision.

When those inputs are reviewed in separate dashboards, the team can agree that a product sold well and still disagree about whether it should be scaled.

Build the SKU margin bridge

Start at order-line level so discounts, quantities and returns can be allocated to the product that created them.

A practical SKU bridge is:

gross product sales
− allocated discounts
− refunds and returns
= net product sales
− product cost
− fulfilment and payment costs
= contribution before acquisition
− allocated acquisition cost
= contribution after acquisition

Show the absolute contribution in pounds as well as the percentage. A high percentage on a small number of orders and a lower percentage on a large number of orders are different decisions.

Shopify’s profit reporting relies on product or variant cost being present at the time of sale. Its documentation also notes that discounts and refunds change the margin reported for the period. Shopify: Profit reports

Five questions that expose a false winner

1. Was the revenue earned at full price?

Discounted revenue can preserve volume while removing a disproportionate share of contribution.

Suppose a product sells for £50 and carries £30 of direct cost before acquisition. At full price, £20 remains. A £10 discount reduces revenue by 20%, but it reduces the £20 contribution before acquisition by 50%.

The point is not that discounting is wrong. The point is that order count and revenue do not show its commercial effect.

2. Did the order stay sold?

Assess returns against the product and cohort that generated them. A sales report cut off at month-end can make a promotion look stronger before its return window matures.

Returns and refunds also need separate logic. Shopify documents that they can occur at different times and that reporting discrepancies can arise when refund and restock dates differ. Shopify: Sales discrepancies

For each material SKU, show:

• ordered units;
• fulfilled units;
• returned units;
• refunded value;
• resellable units recovered;
• return handling or write-off cost where available.

3. Is the product cost current and complete?

A static product cost can become misleading after supplier increases, currency movements, freight changes or a new manufacturing batch.

Assign an owner to the cost file. State whether the number is supplier cost, landed cost or another agreed basis. Treat missing cost as a visible data-quality issue—not a zero.

4. What did it cost to acquire the order?

Platform attribution can help operate campaigns, but it should not silently become the revenue ledger. Keep Shopify net sales and media spend as reconciled facts; show attribution as an assumption used to allocate or interpret acquisition cost.

Where SKU-level allocation is uncertain, use scenarios instead of false precision:

• contribution before acquisition;
• contribution using blended acquisition cost;
• contribution using campaign-attributed acquisition cost;
• break-even CAC for that product or first order.

A useful model makes the assumption visible enough to challenge.

5. What happened after the first order?

Some products are valuable because they introduce customers who buy again. Others look efficient on first-order revenue but attract customers who do not return.

Do not use speculative lifetime value to rescue weak current economics. Start with observed repeat behaviour over a clearly named window—for example 60 or 90 days—and keep the first-order contribution visible beside it.

The product decision view

For the top products receiving meaningful inventory or media investment, bring the following into one table:

Measure	Decision it supports
Net sales and units	Is there real demand after discounts and cancellations?
Contribution before acquisition	Does the product economics work before marketing?
Contribution after acquisition	Can the current acquisition plan scale sustainably?
Return rate and refund value	Is demand staying sold?
Contribution per order	What remains each time the product sells?
Stock cover and lead time	Can the product remain available without overbuying?
Observed repeat rate	Does the product introduce customers who return?

The purpose is not to produce a universal SKU score. It is to make the next action explicit.

Turn the analysis into a decision

Each material SKU should lead to one of a small number of actions:

Scale

Increase media or inventory when demand is repeatable, contribution survives acquisition and the cash requirement is supportable.

Repair

Keep the product but address the constraint: price, discount depth, fulfilment, product cost, return reason, landing-page expectation or campaign mix.

Reposition

Use the product differently—for example as a bundle component, cross-sell or retention offer—when first-order acquisition economics are weak but another role is supported by observed behaviour.

Stop funding

Reduce promotion or replenishment when the SKU repeatedly consumes contribution and working capital without a credible repair path.

What growth and finance should agree

Before the next SKU-level budget decision, write down:

1. The source and definition of net product sales.
2. The product-cost basis and its effective date.
3. Which fulfilment, payment and return costs are included.
4. How acquisition cost is allocated or modelled.
5. The return window used for comparison.
6. Whether repeat behaviour is observed or forecast.
7. The exact decision the view is meant to support.

This short definition document matters more than a polished chart. If the team cannot explain the movement from order to contribution, the ranking will not remain trusted when the recommendation becomes uncomfortable.

The goal is not perfect product attribution

Perfect attribution is not required to make a better SKU decision.

What matters is separating facts from assumptions:

• Shopify records what was ordered, discounted and refunded;
• the maintained cost basis records what the product cost;
• fulfilment and finance inputs record the other commercial charges;
• paid-media platforms provide spend and attribution signals;
• the model shows how those pieces were reconciled.

That gives the team a number it can act on—and a reason it can explain.

Want to see which products actually create contribution?

Neo Analytica builds fixed-scope commercial data foundations for growing UK Shopify brands. We reconcile Shopify sales, paid-media spend, returns, product costs and finance data so growth and finance can evaluate products from the same commercial view.

Book a 20-minute fit call.