A Shopify discount strategy starts with the behavior you want to encourage and the margin you need to keep. A first-order incentive, a loyalty offer, and a bundle promotion solve different problems. Define the commercial limits before writing the email, then verify that the code, product eligibility, and checkout all match what the customer was promised.
Write the rules in plain language
Specify eligible products, customer groups, minimum spend, usage limits, start and end times, and whether other offers can combine. If you cannot explain the offer in a short paragraph, simplify it before making the campaign.
Estimate the contribution left after product cost, discount, fulfillment, payment costs, and expected returns. An order-value increase is not automatically a profit increase. For example, an offer that encourages an extra low-margin item may add handling cost without improving the order enough to justify it.
Compare the incentive with the behaviour you want
Choose the offer after choosing the problem. A first-order incentive aims to reduce hesitation for an eligible new customer. A bundle can make a complete solution easier to buy. A reorder offer may reward a useful repeat behaviour. Applying the same percentage to every customer makes those different jobs harder to evaluate.
Consider an illustrative item priced at $80 with $32 in product cost and $8 in other variable costs. Before a discount, it leaves $40 before fixed costs. A 20% reduction lowers the selling price to $64 and leaves $24 if those costs stay the same. You would need substantially more comparable orders to produce the same contribution, and additional fulfilment or returns could change the calculation again.
That example is not a reason to avoid all discounts. It shows why the target matters. An offer might introduce a profitable bundle, move stock with carrying costs, or acquire a customer who later returns. Write down which outcome you expect and what evidence would support it. Do not justify today's loss with future repeat purchases you have not measured.
Test the offer as a customer
Sendvio supports discount generation, imports, prefixes, synchronization, and checks around eligibility and expiry. Use those tools to organize the offer, then test real checkout scenarios. Include a qualifying basket, a basket below the threshold, an excluded item, and an attempted repeat use where limits apply.
Check the link, displayed code, and any automatic application behavior. If the email promises “15% off selected linen,” the landing page and checkout should use the same scope. Do not bury a limitation in a footer that contradicts the headline. Localized messages need the same practical checks.
Test the boundaries of the promise
Include the cases most likely to generate support questions: a mixed basket with eligible and excluded products, an already discounted item, a returning customer using a first-order code, and a purchase attempted around expiry. Verify the actual configuration rather than assuming the headline describes how the checkout behaves.
For unique codes, check assignment and usage limits. A unique-looking code is not necessarily limited to one customer unless the appropriate restrictions exist. For automatically applied offers, test the destination link and the final basket together. The customer should not be told a code is active when the checkout silently rejects it.
Decide whether to repeat it
Review redemptions, revenue, average order value, and costs together. Compare customers receiving the offer with a suitable baseline where possible, and distinguish a first purchase from a purchase that probably would have happened anyway.
Keep a record of what the offer taught you. If customers only respond when the discount grows, consider improving product explanation, bundles, or service rather than increasing the incentive again. The next campaign should benefit from what you learned about motivation, not simply inherit last month's coupon because it is easy to duplicate.
Separate redemption from incrementality. A customer using a code demonstrates that the code worked in their order; it does not prove they would not have bought without it. Where practical, compare similar eligible groups or evaluate the promotion against a clearly documented baseline, acknowledging differences in season, stock, and audience.
After the campaign, record net revenue under your chosen definition, redemptions, average order value, variable costs, and subsequent returns. Look for unintended behaviour such as splitting orders to reuse an incentive or delaying purchases until the next predictable sale. Repeat the offer when the commercial result and customer experience support it, and change the mechanism when the discount is only making a weak proposition temporarily cheaper.