Free gifts won. Almost nobody is using them.
We sorted 1,138,230 Shopify orders by what the customer actually got. Free gift orders carried 55% bigger baskets and half again as many first-time customers as discounted orders. Merchants spend seven cents of every promo dollar on them.
Two datasets on this page, each labeled where it appears: 24 months of promo spend across 10 established Shopify stores, and per-order anatomy of 1,138,230 cleaned orders across 8 of them. Draft orders, comps, and exchanges are stripped out. Full methodology at the bottom.
Merchants bet 88 cents of every promo dollar on price cuts
Every promotional dollar in the sample, sorted by what it actually was.
Sorted at the allocation level, so an order carrying both a code and a gift is split between the two. Price discounts include codes, automatic discounts, scripts, and post-purchase edits. Seven of the eight order-side stores: one store’s data predates allocation-level splits and is left out of this chart only.
That allocation would be correct if price cuts were the better instrument. Nobody had checked. So we sorted a year of orders by what the shopper was actually handed, and compared them.
Same orders, five kinds of offer, one clear winner
Each order in the sample got exactly one of these: nothing, a price cut, free shipping, a free gift, or a gift plus a price cut. Here is what customers spent in each case.
Net AOV excludes the gift line itself, since it is 100% discounted. These are dollars that hit the merchant’s bank account.
A gift order netted $172.88. A full-price order netted $111.61. A price-discounted order netted $115.65, and that is before you subtract the $24.35 the discount cost to produce it.
The price cut bought about four dollars of basket for twenty-four dollars of margin.
It is not only bigger baskets. It is different customers.
The two cost figures underneath are not comparable, which is the point of the next section. The gift order’s $43 is the gift’s retail price, because Shopify books a free gift as a 100%-off line item; the merchant paid wholesale. The price discount’s $24 is margin, surrendered at face value. Selection caveat, in the open: gift campaigns are threshold-gated, so part of the basket gap is who qualified. That does not explain who showed up.
Price discounts post the worst first-time-customer share on the page: 30.2%, which is below full-price orders at 32.6%. The tool merchants describe as an acquisition lever acquired less than running no promotion at all. Gift orders came in at 46.5%, and gift-plus-discount orders at 54.2%.
Credit where it is due: free-shipping-only orders posted the single best first-time share on the page, 50.1%, at $7.50 an order. Read that as evidence for non-price incentives generally, not for gifts specifically.
That $43 gift did not cost you $43
Shopify books a free gift as a 100%-off line item, so your reports price the gift at retail. That is a sticker, not a cost. You paid wholesale.
A $34 discount costs you $34. A $34 gift costs you $9.
The shopper values both at $34. Only one of them takes $34 out of your margin.
The entire argument, in two sentencesThat spread is why the two "cost" numbers on the chart above are not comparable. The $24.35 on a price-discounted order is margin, surrendered at face value, gone. The $43.23 on a gift order is the gift's retail price, and your real number is whatever your supplier charged you.
What that looks like on a real store
We modeled a standard threshold campaign against 60 days of real order data from a mid-size DTC brand: roughly 8,800 orders at a $197 average order value, offering a free gift that retails at $34 and costs $9, on a $250 threshold. The honest version of this model charges you for every gift, including the 25% of orders that would have cleared $250 anyway.
Monthly result: about 495 orders stretch to hit the threshold, producing $27,225 in incremental revenue against $14,355 in gift cost, for a $4,684 net margin gain after every single gift is paid for.
The same $34 of perceived value, delivered as a discount code, costs $54,230 a month. Identical incentive to the shopper. A $39,875 monthly difference to you.
This one is a model, labeled as a model: real inputs, modeled outcome, no holdout test. Run it on your own numbers with the GWP ROI calculator. It is free and ungated, and it will tell you when a campaign is going to lose money.
Every store with real gift volume shows the same gap
Stores with 1,000 or more gift orders in the 12-month window. Net AOV excludes the gift line, so both dots are money the customer paid.
Three stores in the sample ran enough gift volume to measure: an apparel brand, a golf brand, and a gear brand, in three unrelated categories at three very different price points. All three show gift orders outspending their own full-price orders, by $26 at the low end and $141 at the high end.
The caveat that belongs right here, not in a footer: gift campaigns are threshold-gated, so some of that gap is which orders qualified rather than what the gift caused. Thresholds do not explain the first-time-customer gap, and they do not explain the cost structure.
Gifts are still a November stunt, not a program
Gift spending spikes to 18% of promo dollars in November. Eight of the other eleven months sit under 5%.
Free-gift dollars as a share of each month’s promo spend. One spike at BFCM, a smaller one in April, and low single digits the rest of the time. Compare that to an evergreen discount code, which runs all twelve months.
Merchants already believe gifts work. They reach for them when the stakes are highest and the competition is loudest, which is the single worst month to be measured against everyone else. Then they spend the other eleven months running an evergreen discount code.
The contradiction nobody has noticed: the promotion running twelve months a year is the one with the worst economics, and the one with the best economics gets four weeks.
If you run a gift program always-on, your competition for eleven months of the year is a percentage-off code that the data says is not doing its job.
What the other 88% is actually buying
This report started as a study of discounting, and the discount findings are the reason the gift findings matter. Three of them, briefly.
1. Nobody is targeting anyone
Incidence is the share of each group’s orders carrying any discount. Give-back is dollars per discounted order. Depth is the discount as a share of the basket before it was applied. The two groups are indistinguishable on all three, which is why 67% of discount dollars end up on returning customers: not by design, just because that is who orders.
New customers were discounted on 41.59% of their orders at $23.96 a time. Returning customers: 41.34% at $24.12. That is not a targeting strategy, it is an absence of one. Because returning customers place two thirds of all orders, 67.3% of discount dollars land on people who had already decided to buy from you.
2. Depth past 20% buys nothing
Measured gross of the discount on purpose: depth’s own denominator is this basket, so comparing net baskets would manufacture the decline. Median depth across the sample is 20%, 90th percentile 25%. One residual mechanic worth naming: fixed-dollar codes read as deeper on small carts, which explains part of the fall past 30%.
Median discount depth across the sample is 19.96% of the basket, and 79% of discounted orders sit between 10% and 30%. Five unrelated verticals, no coordination, one number. Pre-discount baskets are flat from 0% through 30% depth and then fall: deeper offers attach to the same carts or smaller ones. Whatever 20% does for your conversion rate, 30% does the same thing minus ten points of margin.
3. The leak is always on, not Black Friday
November takes 14.07% of annual discount dollars against 13.13% of annual gross sales, barely above its fair share. 86% of discount spending happens in the other eleven months. Two structures drive it: 64 evergreen codes that ran 10 or more months out of 12, one of which gave back $349,000 by itself, and a swarm of 132,555 distinct codes of which 99.4% were redeemed exactly once. Merchants re-plan Black Friday every year and never audit the code that quietly outspends it.
We sell a gift app. Here is the case against our own data.
If we buried this section, you would be right not to trust the rest of the page.
Thresholds select for big carts
"Spend $150, get the gift" gates the offer to shoppers who were already close to $150. Some of the basket gap is who qualified, not what the gift caused. This is the strongest argument against the headline number and we have printed it under every chart it touches.
Nothing here is an incrementality test
No holdouts, no lift tests, no ad spend data, because Shopify does not carry any. Every comparison on this page is descriptive: gift orders look like this, discounted orders look like that. If you run a proper holdout test, we would genuinely like to see it.
Gift volume comes from 3 of 8 stores
The 1.14 million orders are real, but only three stores ran enough gift campaigns to measure. That is why we published the per-store breakout instead of a single pooled number, so you can see the effect survives in all three rather than riding on one.
The cost advantage assumes a real wholesale spread
A gift is cheap because you pay cost and the shopper values retail. If your gift is a high-cost item, a gift card, or something with a thin margin, that spread collapses and so does the argument. This is exactly the case where the calculator tells you not to run the campaign.
These are agency clients, not a random sample
Eight established US Shopify brands across five unrelated consumer verticals, all more operationally mature than average. If anything, your promo program is messier than the ones measured here, not tidier.
How to run gifts as a program
Our judgment on top of the data, labeled as judgment.
Set the threshold 20% to 30% above your AOV
Too close and you are paying for orders that would have qualified anyway. Too far and nobody stretches. Your own last-60-days order data gives you the number; you do not need a benchmark for this one.
Pick the gift for its spread, not its price
You want the widest gap between what it retails for and what it costs you. Branded merchandise, samples, and accessories with strong margins beat a discounted hero product every time. Aim for a retail value around 5% to 15% of the threshold.
Run it always-on, not just in November
This is the whole opportunity in one line. The seasonality chart above is a list of eleven months in which almost nobody is competing with you on this mechanic.
Show the progress toward the gift
A threshold nobody can see is a threshold nobody stretches for. The cart needs to say how far away the gift is at every step, or you have built an incentive that only the customers who already qualified will ever notice.
Measure take rate, not revenue
Gift adds divided by eligible orders tells you whether the offer is desirable. Revenue tells you whether the month was good. If take rate is low, the gift is wrong, and no amount of traffic fixes a gift nobody wants.
Two free things that go deeper than this page: the GWP ROI calculator runs the margin math on your own numbers, and the FGWP Playbook walks through gift desirability, tier structure, thresholds, and urgency using one brand's same-month test of two gift campaigns against a 20% sitewide sale.
FAQ
Is gift with purchase better than a discount code?
In this dataset, yes, on every measure available in an order ledger. Free-gift orders netted $172.88 against $111.61 for full-price orders and $115.65 for price-discounted orders. Gift orders reached 46.5% to 54.2% first-time customers against 30.2% for price-discounted orders. And the gift's real cost is what you paid your supplier, while a discount costs face value in margin. The caveat we have to print: gift campaigns are threshold-gated, so part of the basket gap is which orders qualify rather than what the gift caused.
How much does a free gift with purchase actually cost a merchant?
Your Shopify discount report will show the gift's retail price, because Shopify books a free gift as a 100%-off line item. In this sample that came to $43.23 per gift order. Your real cost is what you paid for the gift. On a gift that retails at $34 and costs $9 wholesale, the reported number overstates the true cost by roughly 4x.
What percentage of promotional spending goes to gifts?
7.04%. Across the cleaned 12-month sample, price discounts took 87.98% of promo dollars, free gifts 7.04%, and shipping promotions 4.98%. Gift spending also spikes to 18% of promo dollars in November and sits in the low single digits most other months, which means most merchants who use gifts run them as a seasonal campaign rather than an always-on program.
What should my gift with purchase threshold be?
Set it 20% to 30% above your current average order value. Too close to your AOV and you are gifting orders that would have qualified anyway, which is pure cost. Too far above it and nobody stretches to reach it. Pick a gift whose retail value is roughly 5% to 15% of the threshold, with the widest retail-to-cost spread you can find.
Do discounts bring in new customers?
Not as most stores run them. Across 1.14 million orders, price-discounted orders had a lower first-time-customer share (30.2%) than full-price orders (32.6%). New and returning customers were discounted at nearly identical rates and depths, 41.59% versus 41.34% of orders and $23.96 versus $24.12 per discounted order, so there was no visible targeting at all. 67.3% of discount dollars landed on returning customers.
You sell a gift with purchase app. Why should I trust this data?
You should check it rather than trust it. Every number here is computed by a script from raw Shopify order data, the methodology and the exclusions are published, and the strongest argument against our own finding, threshold selection bias, is printed under the chart it weakens. When cleaning the sample killed one of our early findings, we cut it. The underlying research was published first on our parent agency's site, discount findings and all, and is linked from this page.
What we measured
Sample. 10 established US Shopify stores, seven and eight figures, across five unrelated consumer verticals, all clients of Ethercycle, the agency that builds Promo Party Pro. Longitudinal side: 24 months of monthly gross sales and promo dollars per store from Shopify analytics, July 2024 through June 2026. Order side: 12 months of per-order discount applications on 8 of the 10 stores via the Shopify Admin API, July 2025 through June 2026, totaling 1,138,230 orders after cleaning. Stores are never named.
Cleaning. Excluded: test orders, cancelled orders, draft orders (24,178 manually priced wholesale and B2B orders) and $0-total orders (9,959 comps and exchanges). Those exclusions carried 29% of raw discount dollars and inflated apparent give-back per order by a third, which is why the whole report runs on the cleaned sample. One store's order history begins at its platform migration, so only its post-migration months are included on the order side.
Definitions. A gift order carries a 100%-off automatic line item, which is how Shopify books gift with purchase. Net AOV excludes the gift line itself, so it is the money the customer actually paid. Depth is the discount as a share of the pre-discount product basket, shipping excluded. New versus returning uses the customer's order count as measured at query time rather than order time, so returning shares are modestly overstated.
Limits. The longitudinal promo metric counts codes and automatic discounts only, so those rates are floors. No incrementality claims anywhere. Gift volume comes from 3 of 8 stores. Compare-at markdowns are out of scope entirely: Shopify books a markdown as the price, not as a discount, so no order-level ledger exists for them.
Published August 3, 2026. The underlying research, including the full discount-side findings and 12 charts, was published by Ethercycle as the State of Discounts 2026. Every number on this page is script-computed from the same locked stat manifest.
Run the better promotion
Promo Party Pro is the gift-with-purchase app behind the mechanic in this report: thresholds, tiers, a cart progress bar, and gift stock control, on Shopify checkout.