
How to run email offers that protect your margin
Work out what a discount must earn back, write a discount policy, sell at full price more often and raise order value instead of cutting price.
The Wizel Team
Email marketing strategist · · 9 min read
On this page
- Work out what a discount has to earn back
- Match the offer to the job
- Write a discount policy and keep to it
- Sell at full price most of the time
- Structure a sale so it converts
- Change the offer, not the depth, in a long sale
- Raise order value instead of cutting price
- How to set a discount policy for email
- Frequently asked
Most email calendars run on one offer: 20% off, sent whenever a week looks slow. It works on the day and costs you later, in margin and in customers who learn to wait. This chapter gives you the math for what you can afford to give away, a written policy for when and to whom, and the full-price and order-value tactics that do the job a code was doing.
+50%
more units a 20% discount has to sell to break even at a 60% margin
Work out what a discount has to earn back
A discount comes off profit, not revenue. On a product with a 30% margin, 10% off takes a third of the profit on every order. So before any sale, ask how many extra units the discount has to sell just to stand still. The formula is the discount divided by (margin minus discount).
| Discount | Extra units needed to break even |
|---|---|
| 10% | +20% |
| 20% | +50% |
| 30% | +100% (double) |
| 40% | +200% (triple) |
Revenue is the wrong bar. At a 40% margin, a 20%-off sale that lifts revenue 30% looks like a win and loses money. Three reasons justify a sale that fails the formula anyway: clearing dead stock, buying subscribers whose lifetime value you have verified, and showing up for a moment like Black Friday where absence costs you.
Margin also sets how hard you can compete. At a 70% gross margin a store can bid more than twice what a 30%-margin store can for the same customer, so every point given away out of habit is a point you can't spend winning the next one.
Run the formula before every sale. If you can't argue the discount will sell that many more units, don't run it.
Match the offer to the job
Pick the offer from what you need it to do, never from habit.
| Offer | Best for | Margin cost | Risk |
|---|---|---|---|
| Percent off | A sale event; products under $100 | The full discount on every unit, growing with the basket | Sets a lower reference price and trains waiting |
| Dollar off | Products over $100; high-value welcome offers | A fixed amount, whatever the basket | The same as percent off once it is deep or frequent |
| Bundle price (3 for $105) | Raising order value on consumables, multiples, gifts | A per-unit discount nobody sees | Can replace full-price orders you would have had later |
| Buy one, get one half off | Moving units; clearing a slow product as the second item | 25% off the pair | Reads badly on a considered buy like a watch or a mattress |
| Free shipping over a threshold | Raising order value | Shipping on orders that reach the threshold | Set too low, it pays for orders you'd get anyway |
| Gift with purchase | Premium brands; protecting price | The gift's cost of goods | A cheap gift reads as dead stock |
| Early access, financing, a longer guarantee | Full-price brands; hesitation on big-ticket items | Near zero; financing costs about 2-6% in processor fees | Low, if you honor what you promise |
Two framing rules get more from the same money. Under $100 a percentage reads bigger; over $100 a dollar amount does (“$100 off” beats “25% off” on a $400 product), and it caps your exposure, since $15 off costs $15 however big the cart. And free reads bigger than a percentage: buy two, get one free is a 33% discount that reads as generosity and moves three units.
Write a discount policy and keep to it
A discount policy is one page: which offers, how deep, how often, for whom. Written down, it stops a slow Tuesday from setting your prices, and it stops customers timing you, since they wait for any sale they can predict. The four checks in our guide to discounts that train customers to wait show whether that has started.
- Depth. 10% is a nudge. 15-20% is a standard promotion and your repeatable ceiling. 25-30% is for one or two events a year. 35% and deeper says full price is fiction. If Black Friday is 25%, nothing else all year touches 25%.
- Frequency. Four sitewide sales a year plus Black Friday, at most. At six to nine, customers start timing you, and the tell is a revenue dip in the two weeks before each known sale. Between sales, run scoped offers on one collection or product family.
- Welcome offer. Size it to margin: 15-20% at margins of 70% and up, 10-15% at 50-70%. Below 50%, or above a $300 average order, skip the percentage and offer free shipping, a dollar amount at a spend threshold or a gift. Expire it after 7-14 days.
- Who gets codes. Lapsed and at-risk customers. Not full-price buyers, VIPs or active subscribers, who buy without one.
- Stacking. One discount per order: 20% on top of 20% is 36% off, clearance depth by accident. Free shipping and gifts don't count as discounts.
- Codes and deadlines. Unique single-use codes, which Klaviyo generates in flows and campaigns, so nothing leaks to coupon sites. Set the expiry on the code, not only in the copy. Extend once a year at most, never on the same event twice.
- Was-prices. Only a price you genuinely charged recently. Inflating one breaks pricing rules in the EU, the US and Australia.
If you run email for clients, the policy is the client's pricing decision written down. Agree it per store, because margins differ, and point to it when an approval round asks for a quick 20% off.
Sell at full price most of the time
Most campaigns should give people a reason to buy that isn't a price cut. Outside a real sale window, keep offer-led sends to about 15% of your campaigns, whatever you sell: they buy revenue at the cost of order value, while product-led and story-led sends lift order value and repeat rate.
Hesitation is usually doubt about value, not price. Answer it by teaching something about the category, showing how you make it differently from the norm, sharing customers' specific results and opening up your sourcing. For urgency, use real limits (a small batch, a pre-order window, a seasonal edition, a waitlist) and show the real count. Fake scarcity, caught once, sinks every claim after it. A full-price story runs four emails over 6-8 days, 24-48 hours apart:
- 1.Hook. A story or a surprising fact. No product.
- 2.Value. Teach something and introduce the product softly, with a “learn more” link.
- 3.Offer. The full product, with two or three testimonials that give specific results.
- 4.Close. A real limit: the batch, the deadline, the stock left.
Expect less revenue than a discount send; a rule of thumb is 60-80% of it, so judge on profit and on whether buyers come back at full price. It works best on products of $40 and up with something real to say. A list trained on codes needs 60-90 days of full-price sends before these land, so start with your engaged segment.
Structure a sale so it converts
When you do run a sale, run it as a short sequence with a real end. A sale earns most at the close: as a rule of thumb, 40-60% of its revenue lands in the final 24 hours. Five emails cover it.
| When | What it does | |
|---|---|---|
| Teaser, plain text | 24-48 hours before | Under 100 words. Says something is coming and when, without the discount. Invites people to browse and pick favorites. |
| Launch | The minute the sale starts | Offer, scope, exact end time and button above the fold. Sale prices beside the originals, code applied automatically. |
| Reminder | Day 2 or 3 | Savings in dollars with worked examples, best sellers from the sale, bundles that reach free shipping. |
| Closer | 12 hours before the end | Countdown, end time, button. Often the biggest earner, at 35-40% of sale revenue. |
| Last chance, plain text | 6 hours before the end | Under 75 words, in the founder's voice. |
Send the closer 12 hours out, not 24: a full day leaves too much room to put it off. Then keep the deadline. Extend at most once a year, for email subscribers only, after the public deadline has passed, and never on the same event twice.
Change the offer, not the depth, in a long sale
A sale longer than about five days needs more than one offer. Hold one code for two weeks and three things go wrong: engagement stalls around day 7 because the list has already decided, only the final day carries a real deadline, and nobody buys the same offer twice.
Split the event into windows and give each one a different offer with its own end date. Someone who ignored 20% off may act on a free gift over $100 instead. Each window gets its own last chance, and a day-2 buyer has a new reason to come back on day 9.
- One offer live at a time. Overlapping offers make shoppers wait to see which is better.
- Hold the headline depth roughly level and vary the shape: a bundle, a gift over a threshold, buy one get one half off, a spend-more-save-more ladder.
- Close every window on time, and start the next one instead of extending.
- Keep each offer simple enough to fit in a subject line, and put your strongest one on the peak days.
When a sale stalls, change the offer. Never deepen it: a rising discount teaches everyone to wait for the last window.
Buy one, get one half off is worth knowing: the basket is 25% off, customers see 50%, and it moves two units. It suits consumables, not a $300 espresso machine. If you sell subscriptions, keep most active subscribers out of sitewide sale sends, or they cancel and re-subscribe at the sale rate; send them a discounted one-time add-on for their next delivery instead. The peak season chapter applies all of this to Black Friday.
Raise order value instead of cutting price
A bigger basket is the safest place to spend your generosity. Cross-sells and upsells cost no margin, a threshold costs a little, and a discount costs the most.
- Free shipping threshold. Set it 15-25% above your median basket, not your average: a $70 median points to an $85 threshold. Stock $10-25 add-ons so one item closes the gap, show a progress bar in the cart, and recheck the threshold twice a year.
- Gift with purchase. A gift that retails at $25 and costs you $6 feels as generous as 20% off a $125 order, which costs $25 of margin. Gate it at or above your average order, state its retail value and make it something people want. A sample of a product they haven't bought yet recruits the next full-size order.
- Cross-sells. Three products at most per email or page, chosen for the next problem the purchase creates: the refill for the device, the piece that finishes the outfit. Your order data shows which products customers buy together.
- Bundles. A fixed pack lifts order value and hides the per-unit discount. Pairing a slow seller with a best seller moves stock without marking down your heroes.
Check that a bundle adds value rather than moving it around. Say a product sells for $60 and costs $24 to make, and the customer cost $75 to acquire. Three full-price orders bring $108 of gross profit. A three-pack at 15% off sells for $153 and leaves $81: it covers the $75 at once but earns $27 less. That pays only if the pack adds consumption or brings the customer back; if it replaces orders they would have placed anyway, it cost you.
Start with two numbers: your gross margin and your median basket. Run the formula on your next planned sale, set your free shipping threshold from the median, and write the one-page policy before the next slow week tempts you. Then judge what changed by profit and repeat orders rather than revenue, which is where the next chapter picks up: how to measure email after Apple Mail Privacy Protection.
How to set a discount policy for email
- 1
Find your numbers
Work out the gross margin of the products you plan to discount, and your median basket.
- 2
Run the break-even formula
Divide the discount by (margin minus discount) to see how many more units the sale must sell to stand still.
- 3
Pick the offer from the goal
Use a threshold or bundle to raise order value, a gift or early access to protect margin, and a welcome offer or referral to win first orders.
- 4
Set depth and frequency limits
Keep standard promotions at 15-20%, save 25-30% for one or two events a year, and cap sitewide sales at four a year plus Black Friday.
- 5
Decide who gets codes
Send codes to lapsed and at-risk customers, and keep full-price buyers, VIPs and active subscribers out of mass discount sends.
- 6
Set code rules
Use unique single-use codes, one discount per order, an expiry set on the code, and at most one extension a year.
Frequently asked questions
How often should an online store run a sale?
Four sitewide sales a year plus Black Friday is a sensible ceiling for a brand that wants full-price sales to hold. Between them, run scoped offers on one collection or product family.
Is percent off or dollar off better?
Under $100, a percentage reads bigger; over $100, a dollar amount does. Dollar off also caps what you give away, because the amount stays the same however big the cart gets.
How do I know if a discount will be profitable?
Divide the discount by your margin minus the discount. At a 60% margin, 20% off needs 50% more units just to break even, so a sale that only lifts revenue may still lose money.
Where should I set my free shipping threshold?
About 15-25% above your median basket, close enough that one added item reaches it. Recheck it twice a year as order values change.
