Promotional pricing uses a temporary price incentive on an existing offer for a defined commercial objective and period. This article focuses on promotions that reduce the amount the customer pays.
Evaluate promotional pricing against its stated objective, with incremental contribution as the core economic measure. Revenue growth alone cannot tell you whether the promotion added value. Include campaign costs, effects on other products, and sales after the discount ends.
A useful review asks whether the offer changed buying behavior, what that change earned after costs, and whether the result met the original objective once regular pricing resumed.
Key takeaways
- Set the objective, baseline, spending limit, and review window before launch.
- Allow for the lower contribution on purchases that would have happened anyway.
- Measure related products and later sales so switching and early purchases do not look like new demand.
- Keep uncertain estimates visible when deciding whether to repeat the offer.
Define the promotion before choosing the discount
Start with the behavior you want to change. An established product might need more first-time buyers, renewed interest from inactive customers, or faster movement of aging stock. Each objective requires a different success measure.
For customer acquisition, track the cost of additional buyers and their subsequent contribution. For reactivation, compare purchasing among eligible inactive customers offered the promotion with a comparable group that was not. For inventory reduction, measure stock sold and the resulting contribution; count avoided storage or disposal costs only when you can support them.
Keep the scope clear. A low introductory price for a new offer belongs in penetration pricing. Choosing which items to sell together belongs in bundle pricing. This review concerns a temporary incentive on an offer already available to customers.
Write down these decisions before activation:
- Offer: eligible SKUs, variants, customers, channels, and exclusions.
- Timing: start, end, timezone, and the price that resumes afterward.
- Objective: the primary outcome and minimum acceptable result.
- Budget: additional spend and the maximum contribution loss you will accept.
- Comparison: how you will estimate sales without the promotion.
- Review: when returns, repeat orders, and any sales slowdown will be assessed.
There is no universal best duration. Allow enough time to observe the behavior you are testing, within stock and budget limits. Set stop conditions for an unintended checkout price, a breached spending limit, or fulfillment capacity you cannot meet. Extending the offer changes the test; record it as a decision, not an automatic response to strong sales.
Build a baseline that represents the same trading conditions
The baseline estimates what would have happened during the same period without the offer. Last week’s sales are useful context, but holidays, advertising, availability, and customer mix can make them a poor comparison.
Where practical, randomly assign eligible customers to promotion and control groups before launch. Compare outcomes across everyone assigned, including people who buy nothing. Use contribution per assigned customer when group sizes differ. Comparing coupon users with non-users selects customers by their response to the offer and does not establish its effect.
Decide what the test measures. If the promotion group receives both a discount and extra advertising, the result concerns that combined campaign. To isolate the discount, keep the other conditions as similar as possible. Record code sharing or other exposure that reaches the control group.
If a controlled test is impractical, use comparable periods or a forecast prepared before launch. Check weekday mix, seasonality, stock availability, traffic sources, and other campaigns. Keep the same product variants, market, and currency in the comparison. A nearby substitute is a weak control if the promotion can draw buyers away from it.
Distinguish the measurements:
- Promotional lift is sales above the estimated baseline during the offer.
- Incrementality is the change attributable to the promotion within the scope and period being evaluated.
A promoted SKU can gain units while a substitute loses them. A customer can buy this week instead of next month. Both patterns affect the interpretation of lift.
Keep a record of competing offers and stock changes as context. A suitable competitor price monitoring frequency helps reveal changes during the event, but competitor prices alone do not prove why your sales moved.
Save a lower, central, and upper baseline estimate when uncertainty is material. Do this before seeing the campaign result so the review does not depend on whichever comparison makes the discount look best.
Capture the actual selling price and all additional costs
Use recorded transaction revenue after discounts and refunds. Exclude sales tax collected for remittance, and keep the accounting treatment consistent with the baseline. Decide how shipping income and shipping expense enter the calculation.
Test checkout combinations before launch. On Shopify, product discounts apply before order discounts, with shipping discounts afterward. Multiple percentage order discounts can use the same subtotal rather than reducing a running balance one after another. Eligible combinations depend on discount classes and store settings. Save representative checkout results instead of assuming one stacking formula covers every offer.
The separate list price vs net price guide covers the price arithmetic. For promotion evaluation, the input is the amount actually retained from each sale.
Use the same contribution margin cost policy for the campaign and its baseline. Include product cost, payment fees, fulfillment, and other variable selling costs. Record return handling and unrecovered costs consistently, replacing estimates when returns settle.
Then identify costs caused by running the campaign: extra media, creative work, temporary labor, or another incremental expense. Count each cost once. If shipping support is already included in variable cost per order, it must not appear again as a separate campaign charge.
The discount is already reflected in reduced transaction revenue. Subtracting the advertised saving again would double-count it. Similarly, ordinary fixed overhead that remains unchanged belongs outside the incremental comparison; additional overhead caused by the event belongs inside it.
Calculate incremental contribution: a worked example
Subtract the contribution expected without the offer from the contribution earned during it, then deduct additional campaign costs. This accounts for the lower earnings on baseline purchases as well as earnings from extra units.
Incremental contribution =
Promotion contribution
− Baseline contribution
− Additional campaign costs
Promotion and baseline contribution mean revenue minus variable costs, before the additional campaign costs shown separately. Measure both for the same products, audience, and period. Add related-product and post-promotion effects separately only if they are not already included in that scope.
Consider a hypothetical seven-day promotion for an existing retail SKU. Every figure below is an illustrative assumption, not a store result or industry benchmark. The baseline is 400 units at $80. The promotional price is $68, and 700 units sell.
For simplicity, the example assumes no refunds and $48 of variable cost per unit in both scenarios. Shipping is included in that cost, with no separate shipping income. All prices exclude tax. Additional campaign spend is $600, and ordinary fixed overhead is unchanged.
| Seven-day result | No-promotion baseline | Promotion |
|---|---|---|
| Units sold | 400 | 700 |
| Revenue per unit | $80 | $68 |
| Total revenue | $32,000 | $47,600 |
| Variable cost per unit | $48 | $48 |
| Contribution per unit | $32 | $20 |
| Contribution before campaign spend | $12,800 | $14,000 |
| Additional campaign spend | $0 | $600 |
| Contribution after campaign spend | $12,800 | $13,400 |
Units rise by 75% and revenue by 48.75%. Yet the additional contribution is only $600:
$14,000 − $12,800 − $600 = $600
You can reconcile that result in two parts. The 300 units above baseline contribute $6,000 at $20 each. The 400 units expected to sell anyway earn $12 less each, giving up $4,800. After $600 of campaign spending, $600 remains.
This is an estimate against the assumed baseline. It is neither total business profit nor proof that all 300 extra units represent new demand.
Check the margin for error before approving the offer
Under these assumptions, the campaign must sell 670 units to match the baseline contribution after campaign spend:
($12,800 + $600) ÷ $20 = 670 units
That requires 67.5% more units than the 400-unit baseline. The 700-unit result clears the threshold by just 30 units, before accounting for related products or later sales.
The calculation assumes positive, constant contribution per promotional unit and unchanged campaign spend. It needs revision if fulfillment costs step up or the product mix changes. A required volume is a planning threshold, not a demand forecast.
Baseline uncertainty also matters. If sales without the offer would have been 450 units, baseline contribution would be $14,400. The same campaign would then be $1,000 behind. Test whether the decision survives a reasonable baseline range before treating a small positive estimate as a repeatable win.
Review related products and the period after the promotion
End the discount on schedule, then keep the measurement open. An immediate sales gain can include purchases brought forward, while additional purchases elsewhere in the store can improve the result. Neither effect should be assumed from the promoted SKU alone.
For related products, estimate contribution changes against their own baselines. A customer choosing the promoted item instead of another product can reduce contribution elsewhere: product cannibalization. Additional purchases caused by the promotion can add contribution: a halo effect. Count the incremental change, not every item appearing in a promotional basket.
For later sales, use a follow-up period suited to the normal purchase cycle. A frequently replenished product may reveal a slowdown sooner than a durable item. A drop after the promotion is a signal to investigate, not proof of early purchasing; stockouts, seasonality, and subsequent campaigns can also change sales.
Extend the same hypothetical example. Assume a credible comparison establishes these additional effects. The later period contains no new promotion: 750 units sell at the regular price against an 800-unit baseline. At $32 contribution per unit, the 50-unit shortfall represents $1,600.
| Component of the full review | Change in contribution |
|---|---|
| Promoted SKU during the event, after campaign spend | +$600 |
| Substitute products during the event | −$500 |
| Additional related-product purchases during the event | +$300 |
| Promoted SKU during the follow-up period | −$1,600 |
| Total incremental contribution | −$1,200 |
For this illustration, other effects are assumed to be zero. Each row covers separate products or periods. If your starting calculation already measures the whole store over the entire review window, do not add these adjustments again.
The event that appeared to add $600 now falls $1,200 below the no-promotion scenario. That conclusion depends on the comparison being credible. Without enough evidence to attribute the later shortfall, report it as a sensitivity scenario rather than an established campaign loss.
Separate first-time and established customers
Review these groups separately because their subsequent behavior can differ. Use purchase history before the offer to define established customers, and distinguish genuinely new buyers from returning customers using a different account where your records allow it.
For each group, track net contribution during the event, returns, time to the next purchase, and contribution from later orders. Record whether another discount was needed. Give cohorts the same follow-up time; customers acquired near the end of a campaign have had less opportunity to return at an early review date.
Later purchases from promotion-acquired customers are useful evidence, but they are not automatically incremental. Where possible, preserve the original test comparison through the follow-up window. Keep predicted lifetime value separate from contribution already observed.
Decide whether to repeat, revise, or stop
Return to the objective and spending limit agreed before launch. For an immediate contribution objective, repeat an offer only when the measured gain is credible after costs and the follow-up review. For acquisition or inventory goals, explain any accepted short-term loss in terms of that original objective.
Revise the next test when the result identifies a specific weakness: too much discount on purchases that would happen anyway, an expensive channel, or an eligibility rule that reached the wrong customers. Change a limited number of factors so the next comparison remains useful. Stop when the economics fail the agreed limit or the evidence does not support further spending.
Keep a short decision record with the offer terms, baseline method, actual checkout prices, costs, contribution changes, follow-up cutoff, and unresolved uncertainty. In the example above, higher revenue would not justify repeating the same promotion. A defensible next step would address the measured losses or test whether the baseline was wrong before committing more budget.
See how your own catalog compares with the market
SkuPeek finds matching competitor offers and compares their prices and availability with your products.