Most Shopify gift apps were built with a candle in mind. One product, one variant, one thing to hand over. Point them at a tee shirt and the model starts to creak, because a tee is not one product. It is five sizes and three colours, and fourteen of those fifteen combinations are wrong for any given shopper.
If a gift app makes you choose a single variant up front, you have three bad options. Pick medium and everyone else gets something they cannot wear. Pick a size you are overstocked in and the gift reads as a clearance dump. Pick nothing and skip the offer on the products that actually drive your margin.
Let the shopper choose the size
The fix is not complicated, it is just rarely built. When you add a gift in AOVtools you decide whether to pin it to one variant or leave the variant open. Leave it open and the shopper picks their own size at the moment they claim it, in the cart, before anything ships.
That one decision changes the economics of the whole offer. A free tee someone chose the size of gets worn. A free tee in the wrong size gets returned, which costs you the shipping twice, or gets binned, which costs you the unit and the goodwill you were buying with it.
It also changes what the gift is worth to them. Perceived value collapses the moment a shopper realises the free thing is not for them, and no amount of retail price on the tag brings it back.
Colours, and giving people a real choice
Size is the obvious one. Colour is the one that decides whether the offer feels generous or grudging.
A gift can hold a shortlist rather than a single product. Put three colourways in front of the shopper and let them take the one they want, or three different products if you would rather. They are asked once, in the cart, and the choice takes a tap.
You are still the one approving what is on the list. This is not a store credit where they wander off into your catalogue. It is a small set you picked, priced the way you want it priced, and every option can carry its own settings.
The stock problem apparel has and candles do not
Here is the part that catches apparel brands out, and it has nothing to do with the offer itself.
Your size curve does not sell evenly. Medium and large go first, always. Six weeks into a free gift promotion you are out of the sizes most people want, and the offer either keeps promising something you cannot ship or quietly stops working for most of your customers.
Two settings matter here. The first is a reserve: the number of units you want held back for paying customers. When giving one more away would break into that reserve, the offer steps aside on that product and the product carries on selling at full price. You never end up giving away the last twelve mediums you could have sold.
The second is a backup. Rank a second product behind the first, and when the first is gone or would break the reserve, the next one takes over on its own. Nothing to republish, no gap where shoppers see an offer you cannot honour, and no Monday morning spent noticing.
Giving away the sizes you cannot sell
There is a version of this that works in your favour, and it is worth setting up deliberately rather than by accident.
Every apparel brand ends a season holding sizes at the ends of the curve. The default answer is to mark them down, which puts them next to your good stock at a lower price and teaches people what your product is really worth.
Pin a gift to those sizes instead. The item keeps its price tag, the shopper gets something with a real retail value, and the units leave without a discount attached to your brand. You are trading stock you were going to write down for average order value you were not going to get.
That only works if you pin the variant, which is the opposite of the advice above. Both are right, for different products. The point is that the decision is yours per gift rather than made for you by the app.
Where subscriptions come in
If you sell basics on a subscription, a gift is the cheapest reason to switch that you have. Require a subscription on the gift and it appears only when the shopper chooses the recurring option, at exactly the moment they are deciding.
You can pin it to a specific selling plan too, so the gift comes with the monthly plan and not the annual one, or the other way around if annual is what you are pushing this quarter.
When none of this matters
If your gift is a one size accessory, a tote, a cap, a pack of stickers, most of this is irrelevant and any gift app will do. Pin the variant, set a reserve so you do not run out, and move on.
It starts to matter the moment the gift is a garment, and it matters most on the products with the best margins, which is usually where you want the gift to come from anyway.
Where to start
One gift, one threshold a little above your average order value, variant left open so shoppers pick their size, and a reserve set on the sizes you cannot afford to lose. Run it for two weeks against a normal two weeks and look at average order value and return rate together.
If the size selection is doing its job you will see it in the returns, not just the basket. That is the number most brands forget to check, and it is the one that decides whether a free gift programme is worth running at all.
Start with a single free gift, and move to a tiered ladder once you know your customers respond.
