The gift is not the offer. The threshold is.
We pulled every Q4 order from three Shopify stores across three consecutive Black Fridays. Where the offer named a spend requirement, the average qualifying cart landed at roughly twice the number named, three years running. Where no offer named a number, carts landed about where that store already was.
Three established direct-to-consumer Shopify stores, all clients of Ethercycle, the agency that builds Promo Party Pro, called Brand A, Brand B and Brand C throughout. We disclose the client relationship and withhold the industries: the client list is public, so naming a category would name the company. Windows are October 1 through December 31 in each store's own timezone. Drafts, comps, cancellations and test orders are stripped out. Stores are never named and gift names are paraphrased. Full methodology, and the case against our own numbers, at the bottom.
A named threshold is a target. Shoppers aim at it.
One brand ran the same $50 gift offer in three consecutive fourth quarters. The average qualifying cart landed at roughly twice the threshold every single time.
Label is how far the average qualifying cart landed above the number the merchant picked. The $50 offer ran in all three Q4s and cleared 2x its threshold every time (1.98x / 2.03x / 2.23x), on 8,004 / 9,965 / 717 orders respectively. The 2025 figure rests on the smallest of those: that brand moved most of its gift volume to a differently-named offer that quarter, so read the third year as a much thinner reading than the first two. The multiple is bounded below by 1.0 by construction, since a $50 offer cannot attach to a $40 cart. What is not automatic is where above the line carts settle, and how stable that is year over year.
Three years, three different Novembers, three different assortments, and the number barely moves: 1.98x, then 2.03x, then 2.23x. Over the same three years that brand's full-price average order value was $47.13, $50.02 and $57.01. The gift did not add a few dollars to a typical cart. The offer produced a different cart.
Read the third year with more caution than the first two. That campaign ran on 8,004 orders in 2023 and 9,965 in 2024, but only 717 in 2025, because the brand moved most of its gift volume that quarter to an offer whose name did not state a number. The 2.23x is real and it is thin.
The same brand's $100 tier says something more useful than a second confirmation. It raised the floor and compressed the multiple: $122.91 at 1.23x in 2023, $166.24 at 1.66x in 2024. A higher threshold gets you a bigger cart and a smaller stretch, which is the tradeoff you are actually choosing when you pick the number.
And the contrast, which we did not have to go looking for, because a different brand ran it for us. Across three fourth quarters, that brand never once named a spend requirement in a gift offer.
Q4 2023: three gift campaigns, 3,314 orders, average cart $58.56. That store's own full-price average order value in the same quarter: $58.17.
Same mechanic, same quarter, that store's own shoppers. Forty cents.
The whole argument in one comparisonThat is the shape of the finding. You can argue about how much of a named-threshold cart the gift caused, and we will argue it against ourselves further down. What is harder to argue away is that the same mechanic, run without a number attached to it, moved the basket by less than half a percent.
The honest limit on that comparison, up front: Shopify has no threshold field on an order, so "named a spend requirement" means exactly that, the merchant put the number in the campaign name. Some of the campaigns we count as unnamed almost certainly had thresholds configured that we cannot see. That is why this page compares named against unnamed rather than thresholded against unthresholded, and why the recommendation further down is to name the number rather than to never run without one.
What to do with that: pick a threshold near 2x your current average order value, then look at where carts actually land after a few weeks. If they cluster just above your number, the offer is working as designed. If they cluster at your old average, your threshold is too low to be a target and you are paying for orders you were going to get anyway.
The GWP ROI calculator runs that math on your own numbers, including the orders that would have qualified without the offer. It is free and it will tell you when a campaign loses money.
Gift orders are much bigger. Here is how much of that we can honestly claim.
Every store-quarter that ran a gift program shows gift orders outspending that same store's full-price orders. The size of the gap tracks how the offer was designed, not whether a gift existed.
Net of every discount, and excluding the gift line itself, so both dots are money the customer actually paid. Buy-one-get-one campaigns are excluded from the gift side: they also net a line to $0 but they are a different mechanic. Read this as description, not as lift. A threshold gift can only ever attach to a qualifying basket, and there is no holdout anywhere in this data.
The brand whose offers name a threshold sits at plus 132%, plus 142%, and plus 242% across the three years. The brand whose offers never name one sits at plus 2%, plus 18%, and plus 1%. Same mechanic, same category of merchandise, almost none of the result.
Two things about that third year, since they cut against us. The gap widened to plus 242% while gift volume at that brand fell by two thirds, from 11,124 orders to 3,758, and the campaign mix changed underneath the number: most of 2025's gift orders came from offers that did not exist in 2023 or 2024, attaching to carts around $200 rather than around $100. A rising AOV gap on falling volume and a different mix is not the same finding as three stable years, and we are not going to present it as one.
Read the big numbers as a description of what threshold gift orders look like, not as a measurement of what the gift caused. A "$100 and up" offer can only ever attach to a cart over $100. Part of that gap is the offer selecting big baskets rather than creating them, and this data cannot separate the two.
What the selection objection does not explain is the ordering. If naming a number were doing nothing, the store that never named one would still show a gap, and it does not: plus 1% and plus 2% in two of its three years. That is what "the gift alone does nothing" looks like in an order ledger.
Gifts reach the customers you already had
In all seven store-quarters with a gift program, the first-time-customer share on gift orders was lower than on full-price orders. Not close. Not mixed. All seven.
Seven store-quarters, three unrelated categories, no exceptions: the gift bar is shorter every single time. Caveat that belongs on the chart, not in a footer: the cohort flag is the customer’s order count read today rather than as of the order date, which pushes older orders toward “returning”. The direction survives that; the exact magnitudes for 2023 should not be quoted.
Brand B: 47.5% against 64.4%, then 52.0% against 66.1%, then 54.8% against 68.9%. Brand C: 36.2% against 58.8%, and the same direction in both following years. Brand A's one campaign: 4.5% against 21.1%.
We sell a gift-with-purchase app, so let us be direct about what that means. A gift with purchase is a basket-size lever and a retention lever. It is not an acquisition channel. If you are buying a gift app to bring in new customers, you are buying the wrong tool, and you will conclude it does not work.
There is a plausible mechanism here rather than a mystery: a threshold gift is visible in the cart, and the shoppers most likely to see it, understand it, and stretch for it are the ones who already know what your $80 of product is worth. That is a hypothesis. The order data supports the pattern, not the explanation.
The one-weekend gift campaign does not come back
One brand in this set ran a gift campaign on Thanksgiving 2023 and never ran another one. It is the most instructive store on the page.
Buy-one-get-one is excluded throughout, which is why the yellow bar reads 100%: every one of that brand’s 356 gift orders for the quarter fell inside the BFCM window, and it never ran another campaign across the 140,238 orders it took in the two Q4s that followed. No store-quarter running a year-round program puts more than 51.9% of its gift volume in that window, because a program has the other eleven weeks of the quarter to put volume against.
That campaign was 356 orders in a single day, and it was 100% of the brand's gift-with-purchase volume for the quarter. It gave away $72.81 of retail value on a $116.44 cart, 62.5% of the cart. Then it stopped. Across the next two fourth quarters, 140,238 orders, there was not one gift campaign.
We cannot see why they stopped, and the giveaway ratio invites a conclusion the data does not support. What we can see is the alternative. No store-quarter running a year-round program puts more than 51.9% of its gift volume inside the BFCM window. And the weeks after Cyber Monday are not dead: Brand B landed 36.5%, then 34.5%, then 42.7% of each quarter's gift orders after Cyber Monday, while Brand C's post-Cyber-Monday share climbed to 30.3% as its program matured.
Black Friday is where a gift program is most visible. It is not where one should begin and end. The store that treated it as a weekend event gave away 62.5% of the cart, on one day, and never ran another gift campaign.
The campaigns that ran big were the ones that gave away little
Gift generosity across these campaigns ranges from 5.7% of the cart to 76.1% of it, a thirteenfold spread. Volume does not follow the money.
The five highest-volume gift campaigns here are all cheap gifts. Not one of them gives away more than 18.1% of the cart.
The six highest-volume gift campaigns in the research, plus the most and the least generous, out of 21 that ran 300 or more gift orders. Buy-one-get-one excluded throughout. Gift names are paraphrased to their category and price, because the brands are anonymized and a signature product name would undo that. This is a retail ratio and not a margin one: Shopify order data carries no cost of goods, so what these gifts cost their merchants is not in the dataset. Every one of the five highest-volume campaigns sits at or below 18.1% of cart value, while the most generous gave away 76.1% and ran 302 orders.
Every one of the five highest-volume gift campaigns in this research sits at or below 18.1% of cart value, and the biggest of them, 9,965 orders, gave away a $13 accessory. At the other end, the most generous campaign in the set gave away 76.1% of the cart and ran 302 orders, and the single most generous day belonged to the one-weekend brand above at 62.5% on 356 orders. A $4 sticker on a $61.89 cart carried 1,994 orders by itself.
Perceived value is what moves a cart, and perceived value is not the same thing as cost. That is the entire structural advantage of a gift over a discount: the shopper values the gift at retail and you paid wholesale for it, while a percentage off costs you its face value in margin the moment it applies.
The limit of this chart, stated plainly: these are retail ratios. Shopify order data carries no cost of goods, so nothing here tells you what these gifts cost their merchants or what margin they kept. Treat the percentages as a generosity comparison between campaigns, never as a profit statement.
Most gift campaigns ever run have never been measured
A free gift is never a $0 line item. The product keeps its price and a discount covers it. How that discount gets booked changed in the middle of this dataset, and every analytics tool that keys off the new way reports zero for the old one.
A free gift is never a $0 line item: the item keeps its price and a discount covers it. The natural rule, and the one Shopify’s own Discount Functions produce, is a 100%-off automatic discount. Before 2025 the same campaigns ran on Shopify Scripts, booking a fixed dollar amount equal to the item’s price, so that rule returns nothing for them. In Q4 2023 it finds 0 of one brand’s 3,157 gift orders. Detection here is by outcome instead: a line whose discounts cover its full price, where the covering discount is merchant-run rather than a code the shopper typed. Buy-one-get-one is excluded from both sides; at one store the only thing the 100%-off rule ever found was its BOGO.
The obvious detection rule, and the one Shopify's own Discount Functions produce today, is a 100%-off automatic discount. Before 2025 the same campaigns ran on Shopify Scripts, which booked a fixed dollar amount that happened to equal the item's price. To a report looking for "100% off," that is invisible. In Q4 2023 it finds zero of one brand's 3,157 gift orders, and 92 of the other brand's 11,796.
This one cost us something to publish, because our own earlier research used that rule and undercounted. We fixed it by detecting the outcome instead of the mechanism: a line whose discounts cover its full original price, where the covering discount is merchant-run rather than a code the shopper typed. Buy-one-get-one campaigns net a line to $0 the same way, so they are identified and excluded from every gift number on this page rather than blended in.
If you ran gift campaigns before 2025 and concluded they did not do much, it is worth checking whether your reporting could see them at all.
The practical version: pull your orders, find the line items that netted to zero, and look at what the rest of those carts did. That is the comparison we ran here, and it is the one your dashboard is not running for you.
Do this before you decide gifts do not work for youWe 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.
None of this is an incrementality test
No holdouts, no lift tests, no ad spend, because a Shopify order ledger carries none of it. Every comparison here is descriptive: gift orders looked like this, full-price orders looked like that. If you have run a proper holdout on a gift campaign, we would genuinely like to see it.
Thresholds select for big carts by construction
A "$50 and up" offer cannot appear on a $40 cart, so the multiple is bounded below by 1.0 before anyone stretches for anything. The finding is that carts cluster just above the line and that the multiple holds steady across three years, not that the threshold created every dollar above it.
Three stores is a small sample, and they are our clients
Three brands in three unrelated categories, all clients of the agency that builds this app. We do not name their industries either: the client list is public, so a category plus "our client" is a name with extra steps. That is why every number on this page is broken out per store and per year instead of pooled into one average that could be riding on a single brand's good quarter.
We measure retail value, never cost
Shopify does not carry cost of goods, so a "$20 gift" here is $20 of sticker price. What it cost the merchant is unknown to us. Any margin claim needs your numbers, not ours, which is what the calculator is for.
Thresholds are read from campaign names, and that fails both ways
Shopify has no "threshold" field on an order, so we recovered thresholds by parsing what merchants named their campaigns. A campaign configured with a threshold but named "Free Gift" reads here as having none, so absence of a threshold in our data is not proof there was no threshold. It also fails in the other direction: a gift named after its own price parses as a spend requirement, which produced one bogus 1.94x multiple that looked exactly like corroboration until we caught it. Every comparison on this page is therefore between offers that named a number and offers that did not, which is the thing we can actually observe.
New versus returning drifts on old orders
The cohort flag is a customer's order count read today, not as of the order date, which pushes older orders toward "returning". The direction of the first-time-customer finding is consistent across all seven store-quarters, but do not quote the 2023 magnitudes precisely.
How to run this in Q4
Our judgment on top of the data, labeled as judgment.
Set the threshold first, pick the gift second
The threshold is the part that changes behavior, and it is the only part of the offer you can reason about in advance from your own order data. Start near 2x your current average order value and adjust from where carts actually land.
Put the number in the offer
The store whose gift offers never stated a spend requirement got carts within half a percent of its own average, across three years and every campaign it ran. A threshold the shopper never sees cannot be aimed at. Whatever you configure, say it out loud in the offer.
Go cheap and desirable before you go expensive
Every one of the five highest-volume campaigns in this data gave away 18.1% of the cart or less, and the largest of them was a $13 accessory on a $100 cart. A $4 sticker carried 1,994 orders on its own. Spend the effort on making the gift wanted, not on making it costly.
Make the progress visible in the cart
A threshold nobody can see is not a target. If the cart does not say how far away the gift is at every step, the only shoppers who notice are the ones who already qualified, which is exactly the group you did not need to pay.
Keep it running past Cyber Monday
Brand B landed 35% to 43% of each quarter's gift orders after Cyber Monday, three years in a row. December is the cheapest attention of the quarter and most competing promotional calendars have already gone dark.
Judge it on take rate, not revenue
Gift adds divided by eligible orders tells you whether the offer is wanted. Revenue tells you whether the month was good. This report cannot see take rate, because an order ledger only shows the gifts that were taken, and that is the single number your app should be showing you.
The public market behaves as if it already knows this. Of the nine BFCM 2025 gift campaigns that survived source-checking in our field survey, seven put a number in the offer, and the premium brands staged two: Brilliant Earth, Vuori, Victoria's Secret. That survey is the outside view of this report: same mechanic, named brands, public offers only.
Two free things that go deeper: the FGWP Playbook covers gift desirability, tiers and urgency using one brand's same-month test of two gift campaigns against a 20% sitewide sale, and The State of Gift With Purchase 2026 is the full-year, all-mechanics version of this research across 1.14 million orders.
FAQ
What should my gift threshold be for Black Friday?
Start at roughly 2x your current full-price average order value and watch where carts land. In this data a $50 threshold produced average qualifying carts of $98.83, $101.68 and $111.72 in three consecutive Q4s against full-price averages of $47.13, $50.02 and $57.01. A $100 threshold at the same brand produced bigger carts but a smaller stretch, 1.23x and 1.66x. Higher threshold, higher floor, less lift per order.
Does a free gift with purchase actually increase average order value?
Gift orders in this data were much larger than full-price orders, by 132% to 242% at the brand whose offers named a threshold. We will not call that lift, because a threshold offer can only attach to a qualifying cart, and there is no holdout in the data to separate selection from behavior. The cleaner evidence is the contrast: at the store where no gift offer ever named a spend requirement, gift orders averaged $58.56 against that store's own $58.17 full-price average. Naming the number is doing the work.
Should I run a gift with no minimum spend?
If your goal is basket size, attach a number to it and say so in the offer. One store here ran gift campaigns across three fourth quarters and never named a spend requirement in any of them; its gift orders came in within half a percent of its own full-price average, which means the merchandise went out on orders that were already going to happen. If your goal is goodwill on every order, that is a real goal and it is not what this page measures.
Do gift campaigns bring in new customers?
Not in this dataset. In all seven store-quarters with a gift program, first-time-customer share was lower on gift orders than on full-price orders, by as much as 17 percentage points in 2025 alone. Treat gift with purchase as a basket-size and retention lever. If acquisition is the job, this is the wrong tool and you will conclude it does not work.
Should a gift campaign only run over Black Friday weekend?
The brands running gifts as a year-round program never put more than 51.9% of a quarter's gift volume inside the Thanksgiving-to-Cyber-Monday window, and the longest-running of them landed 35% to 43% of each quarter's gift orders after Cyber Monday, three years in a row. The one brand here that treated a gift as a single-weekend event put 100% of its gift volume in that window, gave away 62.5% of cart value on one day in 2023, and never ran another gift campaign.
Why can't I find gift campaigns in my own Shopify reports?
Because of how the discount is booked. Shopify's Discount Functions record a gift as a 100%-off automatic discount, but before 2025 the same campaigns ran on Shopify Scripts, which recorded a fixed dollar amount equal to the item's price. Any report looking for "100% off" returns nothing for those. In Q4 2023 that rule finds zero of one brand's 3,157 gift orders. Look instead for line items whose discounts cover their full price.
You sell a gift app. Why should I trust this?
Check it rather than trust it. Every number is computed by a script from raw Shopify order data, the exclusions and the detection rule are published above, and the finding that hurts us most, that gifts reach returning customers rather than new ones, is a full section rather than a footnote. We also published the fact that our own previous research undercounted gift orders by using the wrong detection rule.
What we measured
Sample. Three established US direct-to-consumer Shopify stores in three unrelated categories, all clients of Ethercycle, the agency that builds Promo Party Pro, called Brand A, Brand B and Brand C throughout. The categories are withheld on purpose. Ethercycle's client list is public, so an industry label plus "our client" identifies the company in one step, which makes the industry a name rather than a description. Every order placed in Q4 of 2023, 2024 and 2025, pulled through the Shopify Admin API with line items retained, 353,466 orders exported. Windows are October 1 through December 31 in each store's own timezone, which matters: a naive UTC window would have misfiled between 17 and 73 orders per store-year.
Cleaning. 343,394 orders analyzed. The 10,072 removed are fully accounted for: 5,089 cancelled, 4,339 draft (manually priced wholesale and B2B), 644 zero-total (comps and exchanges), and no test orders. That exclusion is not cosmetic. One store's Q4 2023 draft orders alone carried $499,006.71 of discount dollars, which would have swamped every ratio on this page. Shipping discounts are stripped out of discount totals throughout, because free shipping is neither a price cut nor a gift.
How a gift is identified. By outcome, not mechanism: a line item whose discount allocations cover its full original price, where at least one covering discount is merchant-run (an automatic discount, a Shopify Script, or a discount function) rather than a code the customer typed. Buy-one-get-one campaigns also net a line to $0 and are identified and excluded from every gift figure here. Free items redeemed with a customer-entered code are excluded too. That "merchant-run" requirement earns its keep: one store-quarter contained 232 lines that netted to zero and zero merchant-run discounts of any kind, the largest of them a damaged-item replacement. A rule that only asked "did this line net to zero" would have reported a gift campaign where there was a returns desk.
Three further exclusions, each of which changed a number on this page. Free gift cards are not gifts: store credit is redeemable for money, so it belongs on the discount side by the same logic that strips free shipping, and 482 such lines came out of one store-year where they had been inflating both the giveaway value and the cart it rode in on. Campaign rows whose gift retail exceeds the cart it attached to, or whose qualifying cart is $0, are bundle-builder and loyalty mechanics rather than gifts, and four such rows are removed. Campaigns where the merchant listed the gift SKU at a nominal price, typically a penny, are excluded from every value figure while their order counts are kept.
Definitions. AOV is net subtotal divided by orders, so it is money the customer actually paid and it excludes the gift line itself. Thresholds are recovered by parsing campaign titles, because Shopify carries no threshold field on an order, and a dollar figure counts as a threshold only where the name also carries a spend signal such as "over" or "+". New versus returning uses the customer's order count as read at query time rather than at order time. The BFCM window is Thanksgiving through Cyber Monday in each year, store-local.
Verification. Every figure on this page is generated by script from a single locked stat manifest; none of it is typed by hand. Nine store-months of the 2025 data reconcile to the cent against an independently built order cache from earlier research, every order's line-level gift allocations sum to its order-level totals with zero mismatches, and two store-years match Shopify's own order counts exactly.
Limits. No incrementality claims anywhere. Take rate is not measurable from an order ledger, which only records gifts that were taken and never the qualifying carts that declined one. Gift cost and gift margin are not in Shopify's order data at all. One further store was in scope for this analysis and is absent for a boring reason: its stored API authorization lacked order scopes at pull time.
Published August 11, 2026. Companion to The State of Gift With Purchase 2026, which covers a full year and all promotion types across a larger store set, and to the State of Discounts 2026 published by Ethercycle.
Set a threshold. Watch the carts move.
Promo Party Pro runs the mechanic in this report on Shopify checkout: spend thresholds, tiers, a cart progress bar so the target is visible, variant picking, and automatic pause when the gift runs out.