A campaign can produce plenty of discounted sales and still be a poor deal for your store. Some customers would have bought at full price. To decide whether to keep an offer, measure the path from seeing it to using it, then check what those orders contributed after the discount.
Start with five useful numbers
| Metric | What it tells you | What it does not tell you |
|---|---|---|
| Views | How often the offer was shown | Whether every viewer noticed it |
| Claims or clips | How often shoppers acted on the offer | Whether they later bought |
| Redemptions | How many times the discount was used | Whether the order was incremental |
| Attributed sales | Order value associated with use of the offer | Profit or causal lift |
| Contribution | What remains after product costs, discount and variable costs | Long-term customer value |
Perk's Analytics page shows daily views, clips, redemptions, and sales, with date ranges and a per-campaign breakdown. Use those numbers as a starting point. Keep your cost and margin calculation in your own reporting, because sales are not profit.
Read the funnel before changing the discount
Divide clips by views to get a clip-through rate. If views are healthy but clips are low, the offer may be unclear, poorly placed, or simply unappealing for that product. If clips are strong but redemptions are low, inspect the checkout path, product eligibility, and the final price. A larger percentage is not always the right fix.
Compare like with like. A weekend sale and a quiet weekday have different traffic. Product mix, stock, ad spend, and seasonality can all change results. Choose a comparable period and make one meaningful campaign change at a time so you can understand what moved.
Calculate what the campaign actually leaves behind
For each discounted order, start with the amount paid for products. Then subtract product cost, the cost of fulfilling the order, payment fees, and any shipping subsidy or returns allowance. The exact accounting depends on your store, but it should be consistent across campaigns. A higher basket value helps only when enough of that value remains.
Do not call all attributed sales "extra sales." If two offers can apply to one order, campaign-level attribution may show that order under both offers. Perk's overall sales view counts the order once, while the per-campaign view attributes its value to each redeemed campaign. That makes campaign comparisons useful, but it is not a causal test.
Decide what to do next
- Keep it: the offer reaches shoppers, is used, and meets your contribution target.
- Improve it: diagnose the weakest funnel step before increasing the discount.
- Pause it: it mostly discounts orders that appear likely to happen anyway or falls below your margin floor.
Record the campaign goal and the minimum acceptable contribution before launch. That makes the eventual decision easier than judging a chart after seeing a large revenue number.
Perk tracks offer views, clips, redemptions, and attributed sales by campaign so you can see where shoppers drop out.
Explore Perk on ShopifyFrequently asked questions
What should I measure for a Shopify discount campaign?
Track who saw the offer, who claimed it, how many orders used it, the sales on those orders, and the margin after the discount and other variable costs. Compare results with a similar prior period or another product, and avoid treating attributed sales as proof that the discount caused every order.
Is revenue from discounted orders the same as incremental revenue?
No. Some shoppers would have bought without the offer. Discounted-order revenue describes orders associated with a campaign; incremental revenue requires a credible comparison, such as a controlled test or a carefully matched baseline.