Gift with purchase: planning the free units you give away

The most expensive sentence in promotion planning is "it's free, so it doesn't touch the numbers." A gift-with-purchase unit is real stock: it was bought at cost, it depletes a real SKU, and it competes with paid demand for the same shelf. Plan it as marketing with no inventory consequences and you get the two classic failures: the promo runs out of gifts in week two, or it ends cleanly and next quarter's reorders are inflated by a spike that was never organic. This guide covers the middle ground where GWP actually gets managed: sizing the free-unit quantity with worked arithmetic, the demand distortions to expect during the window, and the reorder-math cleanup afterward.

One boundary first, because the confusion is common: units that never touch a sale — rep samples, PR pulls, demo units — are never-sell stock with their own parking discipline, covered in the sample inventory guide. Gift-with-purchase units are the opposite: each one exists because a sale happened. They're a selling cost attached to demand, which means they belong inside your demand planning, not parked outside it. Same shelf, same SKU, completely different bookkeeping.

Three shapes of giveaway, three planning problems

GWP gets the deep treatment below because its quantity is the only one you have to forecast. The other two are budgeting exercises (fixed units, known cost) followed by the same cleanup afterward. BOGO sits outside this table on purpose: its free units arrive only in pairs, which changes the per-unit math rather than the pool sizing — the BOGO inventory planning guide owns that arithmetic. Baby and kids stores run this math on a clock of their own — registry signups and shower seasons turn giveaway demand spike-shaped; the baby and kids store inventory guide plans that calendar.

Sizing the GWP pool: the worked example

Illustrative setup: a coffee-bean retailer runs a three-week "free enamel pin with any order over $60" promo. The gift is PIN-EN-050 at a $2.10 unit cost (illustrative).

• Baseline orders. The store averages 3.2 orders a day over the last 90 days.
• Qualifying share. The $60 threshold catches about 55% of orders (illustrative); pull your own order-value distribution, which is one filter in Shopify's analytics (Shopify Help Center, accessed November 2026).
• Uplift assumption. Promotions lift order rate. Assume +30% during the window, an illustrative planning assumption rather than a benchmark. Set it from your last promo's actuals, or from nothing at all for a first promo, and let the buffer absorb the error.
• Expected gifts. 3.2 × 1.30 ≈ 4.2 orders/day; 4.2 × 0.55 ≈ 2.3 qualifying orders/day; 2.3 × 21 days ≈ 48 gifts.
• Buffer. Add 15% for staff hand-outs on borderline orders, service-recovery gifts, and the attach rate running hot: 48 × 1.15 ≈ 56 units to set aside.
• Cost. 56 × $2.10 = $117.60 of promo cost (illustrative). Budget it like ad spend, because it is.

Then the stock check, the step that gets skipped: the store holds 40 pins, and the pin also sells organically at about 0.4 a day. Over the 21-day window it would sell roughly 8 units organically (0.4 × 21 ≈ 8). The promo needs 56 more on top of that: 56 + 8 − 40 = 24 units to reorder before launch. The arithmetic enforces the rule it exists for: never gift your bestseller. A gift SKU that also carries real organic demand will fight the promo for the same units, the promo usually wins, and paid buyers of that SKU face an empty mid-window shelf.

What the window does to your sales data

Three distortions, all temporary, all predictable enough to plan around:

• The promoted SKUs' velocity is promo-flavored. The coffee sells faster because of the pin. Any reorder math that recomputes on this window inherits the spike and carries it forward for months.
• The gift SKU drains by two paths at once. Organic sales and giveaways both deplete PIN-EN-050, and the system's depletion doesn't distinguish them; only your ledger does. That's why the gift count and buffer get written down before launch, not reconstructed after.
• Order value clusters at the threshold. Baskets bunch just above $60 and just below it, where the gift isn't worth chasing. Promo-week average order value compared against baseline tells you about threshold design, not about customers; don't carry that comparison into next quarter's planning.

The spike mechanics are the same as any short promotion; the echo pattern is covered in detail in the flash-sale inventory planning guide.

The reorder-math cleanup afterward

Here's where uncorrected promos cost real money. Continuing the worked example:

During the window, PIN-EN-050 depletes at roughly 0.4 organic + 2.3 gifts ≈ 2.7 units a day. A naive velocity average that treats all 13 recent weeks equally:

(10 weeks × 0.4 × 7) + (3 weeks × 2.7 × 7) = 28 + 56.7 = 84.7 units ÷ 13 weeks ≈ 6.5 units/week

against the true organic rate of 0.4 × 7 = 2.8 units a week. The apparent rate is more than double reality, and a reorder point built on it orders roughly twice what organic demand justifies. That's how a three-week promo becomes four months of slow pins.

Two fixes, and use both:

• Exclude or baseline-correct the window. Drop the promo weeks from velocity calculations, or substitute baseline weeks, before recomputing reorder points. Building a baseline that survives promotional noise is covered in the baseline demand guide.
• Book the depletion as promo expense, not shrink. The 56 gifted units left inventory with no revenue line. Record them (56 × $2.10 = $117.60, illustrative) against the promotion's cost with a cause code in your adjustments. Gift depletion that lands in shrink variance becomes a mystery you re-investigate at every count.

Event giveaways and influencer seeding

Fixed quantities, lighter math, same cleanup rules:

• Event giveaways. 200 seed packets for a garden fair is a budget line, not a forecast. Take a pre-counted case, log what comes home, and keep the giveaway units out of the sellable pool for the event window; the parking mechanics are the same ones used for samples, already covered in the sample inventory guide. The only GWP-style risk is over-gifting on the day, which the pre-count makes visible.
• Influencer seeding. Product sent to creators is fixed-cost marketing with spread-out timing. Track units and cost per creator so "seeding" is a number rather than a drawer. Two compliance notes: creators receiving product generally must disclose the material connection under the FTC's endorsement guides (FTC Endorsement Guides, accessed November 2026), and business gifts to individuals brush against the IRS's $25-per-recipient annual cap in some cases, which is worth an accountant's confirmation before deducting the program (IRS Publication 535, accessed November 2026).

One mechanics pointer, since this article is about the numbers: how you configure the trigger in Shopify (automatic discounts, bundles, or an app) is setup work, covered in the Shopify bundle inventory guide. Nothing in those mechanics changes the arithmetic above.

The GWP planning worksheet

Copy this into the promo doc and fill the middle column; the right column is the coffee retailer's example:

The last row is the one everyone skips. Set the date you'll re-run velocity without the promo window before the promo starts — cleanups scheduled after the fact don't happen.

Common questions

Does gift-with-purchase inventory count as real inventory? Yes. The units were bought at cost, deplete real SKUs, and compete with paid demand for the same stock. They're a selling cost attached to sales: plan the quantity before the window opens, book the depletion as promo expense, and never gift a SKU your organic demand depends on.

How many free gift units should I stock for a promotion? Multiply baseline daily orders × your uplift assumption × the share of orders that qualify at your threshold × promo days, then add a 15% buffer and check the result against stock on hand plus organic sales of the gift SKU. The worked example above walks every step with numbers.

Is a gift with purchase the same as a sample? No. Samples — rep units, PR pulls, demo stock — never touch a sale and belong outside the sellable pool entirely. A GWP unit exists because a sale happened, so it stays inside demand planning as a selling cost. Same physical shelf, opposite bookkeeping and opposite planning.

Does a GWP promotion inflate my reorder math? It does if uncorrected. The window inflates velocity for both promoted SKUs and the gift SKU; in the worked example, the pin's apparent rate more than doubled. Exclude or baseline-correct the promo weeks before recomputing reorder points, and log gift depletion with its own cause code so it never reads as shrink.

Can I deduct influencer product seeding? As a general rule, yes — deducted at cost as an ordinary marketing expense. Two edges to check: the IRS caps deductible business gifts at $25 per recipient per year, which some seeding programs brush against, and treatment varies with how the product is used (IRS Publication 535, accessed November 2026). Confirm with your accountant before deducting the whole program.

Next step

One number decides the whole worksheet: the share of your orders that clears the threshold. Fetch it from your last 90 days of orders today, then fill in the worksheet above for one gift SKU: baseline orders, qualifying share, uplift, buffer, cost, stock check, cleanup date. If the stock check says reorder before launch, add the gift SKU to this Monday's buying list — and the baseline method that keeps post-promo reorders grounded is covered in the guide to baseline demand.