Shopify Discount Statistics 2026: What 1.14 Million Orders Reveal About Where Your Promo Budget Actually Goes

Introduction

If you run discounts on a Shopify store, you already know the feeling: you check your margin at the end of the month and it's thinner than the sales calendar should have caused. New Shopify discount statistics put a number on that feeling, and it's not the number most merchants expect.

Ethercycle, a Shopify-focused agency, published "State of Discounts 2026" using first-party order data from 10 established US Shopify stores across 24 months ($285.3 million in gross sales, $29.4 million in discounts) plus a deeper 12-month, 1.14-million-order sample from 8 of those stores. It's one of the largest first-party looks at real Shopify discount behavior published this year, and several of its findings cut against the standard BFCM-first playbook.

This isn't a report about whether to discount. It's a report about where discount dollars actually go once a merchant hits "publish" on a discount code, and the answer is: mostly not where you'd guess. In this article you'll get the headline numbers, the two findings most worth acting on this quarter (an evergreen-code leak most merchants don't know they have, and a customer-acquisition result that inverts conventional wisdom), and a short framework for auditing your own store's discount data using tools you likely already have.

The stores in this dataset skew larger (seven and eight figures in revenue), but the mechanics it uncovers aren't scale-dependent. A code with no end date leaks the same way at $800K in annual revenue as it does at $10M; if anything, smaller teams are less likely to have a dedicated person tracking code lifecycle, so the leak often runs longer before anyone notices.

A note on the source, upfront: Ethercycle discloses that its team also builds Promo Party, a gift-with-purchase app for Shopify. That's a real conflict of interest on the findings that favor gift-with-purchase over price discounts, and we're flagging it the same way we'd flag it for any vendor-authored study. The underlying order data and methodology are still worth engaging with; just weigh the gift-with-purchase conclusions with that disclosure in mind.

The Data Behind These Shopify Discount Statistics

Before the findings, the methodology matters because it's unusually transparent for an industry report:

• 10 stores, seven- and eight-figure revenue, agency clients, 24 months of order history (July 2024 to June 2026)
• $285.3 million in gross sales, $29.4 million in discount dollars
• A deeper slice: 1.14 million individual orders across 8 of the 10 stores over 12 months (July 2025 to June 2026)
• Pulled via the Shopify Admin API and each store's own analytics dashboards, no customer-level PII collected
• Excluded draft orders, cancellations, and $0 transactions, which the report says represented roughly 30% of raw discount dollars if left in
• The report explicitly states it makes no incrementality claims: there are no holdout groups and no lift tests, so the new-customer and AOV figures below are descriptive comparisons, not causal proof

That last point matters for how you should read every stat below: this is what happened across these orders, not a controlled experiment proving discounts caused or failed to cause an outcome. Treat the findings as a strong prior to test against your own numbers, not a verdict. One more note before the findings: every dollar figure in this article is reported in USD as general commentary on the dataset, not tied to any specific merchant's currency.

Finding 1: 86% of Discount Spending Happens Outside November

The BFCM-first mental model says most of a Shopify store's annual discount budget goes out the door in a four-day window in late November. The data says the opposite. November accounted for just 14.07% of the year's discount dollars, against 13.13% of annual gross sales, meaning the month is only mildly overweighted relative to its size, not the dominant discount event the planning calendar treats it as.

86% of discount dollars are spent outside November. Most of that isn't a deliberate spring or summer campaign; it's leakage from codes that never expire.

The report's code-level breakdown makes the mechanism visible:

• Distinct discount codes redeemed (12 months): 131,912
• Codes used exactly once: 99.4% of all distinct codes
• Dollar share of single-use codes: 57.27% of total code discount dollars
• Codes active 10+ months ("evergreen"): 59 codes
• Dollar share of evergreen codes: 11.62% of total code discount dollars
• Largest single evergreen code: $349,000 in discounts over 12 months

Two different problems are hiding in those numbers. First, a huge number of one-off codes (influencer codes, customer-service goodwill codes, one-time partnership codes) are individually small but collectively carry more than half of all code-based discount spend, which suggests weak tracking and cleanup discipline more than it suggests a deliberate strategy. Second, a small number of codes that were meant to be time-boxed (a launch promo, a partnership code, an "early supporter" discount) never got shut off, and the single largest example quietly cost a store $349,000 over a year it was live.

Neither problem shows up on a BFCM planning spreadsheet, because neither one is a campaign. They're configuration debt.

Finding 2: Discounted Orders Don't Reliably Bring In New Customers

This is the finding most worth sitting with if your team justifies discounting as an acquisition tool. Across the report's order-level sample:

• Price-discounted orders had a lower new-customer share (30.2%) than full-price orders (32.6%).
• Returning customers were discounted on 38.4% of their orders; new customers were discounted on 35.4% of theirs.
• 67.6% of all discount dollars went to customers who had already purchased before, the shoppers with the highest likelihood of buying anyway.

In other words: on average across this sample, a customer who paid full price was slightly more likely to be new than one who received a discount, and the majority of the discount budget subsidized people the store didn't need to convince. This tracks with a pattern retailers have quietly known for years but rarely quantify: loyalty and win-back flows over-index on your best, most loyal segment because they're the easiest to target, not because they're the segment most in need of a nudge.

The report is careful to note this isn't causal, since there's no holdout test proving the discount had zero effect on any individual purchase decision. But as a directional signal, it argues for treating "will this discount bring in someone new" as a question to check per campaign, not an assumption to bake into every promo brief.

A caveat worth flagging here, and this is PromoOS's inference, not something the report itself breaks out: the 30.2% figure above is an aggregate across every discount type in the sample, and the report doesn't separate performance by discount mechanism. It's plausible some types skew that blended number in either direction. Referral discounts in particular seem like a likely candidate to outperform on new-customer share specifically, since they're typically structured around bringing in someone new (issued to an existing customer for referring one) rather than applied broadly to anyone who finds a code. We haven't seen data confirming that gap for referral discounts, and the report doesn't test it, so treat this as a hypothesis worth checking against your own program's numbers, not a finding.

Finding 3: Gift-With-Purchase Outperformed Price Discounts on the Metrics That Matter Most, But Got a Fraction of the Budget

This is the finding where the report's authorship matters most, so read the numbers, then read the caveat. Net AOV, new-customer share, and promo cost per order by promo type:

• Full price (no promo): net AOV $111.61, 32.6% new-customer share, $0 promo cost per order
• Price discount: net AOV $118.40, 30.2% new-customer share, $24.26 promo cost per order
• Free shipping only: net AOV $160.00, 50.1% new-customer share, $7.48 promo cost per order
• Gift with purchase (gift only): net AOV $172.88, 46.5% new-customer share, $41.00 promo cost per order (COGS)
• Gift plus discount combined: net AOV $191.00, 54.2% new-customer share, $80.00 promo cost per order

Gift-based and free-shipping promotions correlated with meaningfully higher new-customer share and higher net AOV than straight price discounts in this sample. Despite that, price discounts absorbed 87.98% of all promo dollars in the dataset, versus 7.04% for free gifts and 4.98% for free shipping.

The report itself discloses a real limitation here: gift-with-purchase offers are usually gated behind a spend threshold ("spend $150, get a gift free"), which selects for customers who were already going to build a bigger cart. That's a form of selection bias the authors call out directly rather than hide, and it's worth taking seriously, since threshold-gating alone could explain a meaningful chunk of the AOV gap.

Given that Ethercycle's team builds a gift-with-purchase app, the incentive to find exactly this result is obvious, and we'd flag the same conflict if a discount-code app published a study concluding price discounts win. The directional pattern (non-price promotions correlating with more new customers per dollar spent) is plausible and worth testing on your own store. Treat the exact dollar figures as one vendor-run study's numbers, not an industry consensus.

Finding 4: Discount Depth Has a Ceiling Merchants Are Blowing Past

The median discount across the dataset was 19.98%, and 84% of all discounts clustered between 10% and 30% off. Cart size held flat across that entire 10-30% range: a 12% discount and a 28% discount produced comparably sized carts. Beyond 30% off, cart size started to decline.

The practical read: for most stores in this sample, there was no incremental basket-building benefit to discounting past the 30% mark, and stacking to 40% or 50% off traded margin for a smaller cart, not a bigger one. If your promo calendar treats "deeper is better" as a default lever for a slow week, this is the finding that argues against it.

Finding 5: 7 of 10 Stores Are Already Cutting Their Discount Rate

Maybe the most quietly important number in the whole report: 7 of the 10 stores voluntarily reduced their discount rate year over year, and the sample-wide average fell from 11.0% of gross sales in year one to 9.8% in year two. Nobody made them do this. Absent a mandate, most of the merchants in this dataset pulled back on their own, which is consistent with operators noticing the same margin drag this report quantifies and self-correcting before they had the data to prove it.

A 4-Point Framework for Auditing Your Own Discount Spend

None of what follows is in the Ethercycle report; it's PromoOS's own checklist for applying its findings to your store.

You don't need a 1.14-million-order dataset to check whether your store has the same leaks. Most of these checks take under an hour using your existing Shopify admin and order exports.

1. Map every active discount code against a calendar, not a list. Pull every code that's currently enabled in Shopify and plot its start and end date. Any code with no end date, or an end date more than 90 days out, goes on a review list. This is the single fastest way to find your version of the $349K evergreen leak.
2. Split your last 90 days of orders into new vs. returning, discounted vs. full price. A basic Shopify Analytics or order export cut will get you close. If your discounted orders skew more toward returning customers than your full-price orders do, your acquisition-focused promos may be doing loyalty work instead.
3. Chart your discount depth against average order value for the last two quarters. Look for the point where deeper discounts stop growing the cart. That's your ceiling, not 30% by default; every catalog and price point is different.
4. Count your one-time-use codes and total their dollar value. If a large share of your discount spend sits in codes used exactly once (customer service saves, influencer one-offs), that's a signal to tighten issuance and tracking, not necessarily to stop issuing them.

The PromoOS Perspective

None of the findings above are really about whether to discount. They're about visibility and expiry: whether a merchant can see, in one place, every discount currently live, when it's scheduled to end, and whether it's colliding with something else.

That's an operational gap more than a strategy gap. A merchant running promotions through Shopify's native discount screen one at a time has no single view of every code's start and end date, so an evergreen code surviving 10+ months isn't a decision anyone made; it's a code nobody remembered to look at. PromoOS's calendar view exists specifically to make every scheduled promo, its start date, and its end date visible in one place, with automated deployment and revert so a promo actually turns off on the date it was supposed to instead of quietly running forever. Conflict detection flags when two live promos would overlap or double-discount the same products before either one ships, and the analytics dashboard tracks revenue and order counts per campaign so a merchant can check the new-customer question from Finding 2 against their own numbers, not just the report's.

The goal isn't to run fewer promotions. It's to make sure every promotion a merchant runs is one they'd choose to be running if they looked at it today, not one that's still live because nobody checked.

Key Takeaways

• Only 14% of a Shopify store's annual discount spend, on average in this dataset, is tied to November: the other 86% is spread across the rest of the year, much of it through codes with no expiry date.
• Discounted orders in this sample brought in new customers at a slightly lower rate (30.2%) than full-price orders (32.6%); most discount dollars went to repeat customers already likely to buy.
• Discount depth showed no incremental cart-size benefit past roughly 30% off; deeper discounts mainly cost margin without growing baskets.
• Gift-with-purchase and free-shipping promotions correlated with higher AOV and new-customer share than price discounts in this data, but the report's authors also build a gift-with-purchase app, so weigh that finding with the disclosed conflict in mind.
• 7 of 10 stores in the study cut their discount rate year over year without being told to, a signal worth taking seriously even without a controlled experiment behind it.

Conclusion

The headline lesson from these Shopify discount statistics isn't that merchants discount too much or too little. It's that most of the damage in this dataset happened quietly, outside the big campaigns merchants actually plan for: evergreen codes nobody remembered to end, and discount budget defaulting to customers who didn't need convincing. Both are visibility problems before they're strategy problems.

The one action you can take today: pull the list of every discount code currently active in your Shopify admin and check the end date on each one. If a code has no end date, or one that's months in the past its intended window, that's your first fix.