Gift vs Discount vs Free Shipping: Ranked by What They Really Cost

Gift vs Discount vs Free Shipping: Ranked by What They Really Cost

Three offers can all lift conversion: a free gift, a percentage off, and free shipping. Most stores pick by gut. The right way is to rank them by cost per incremental order — and the differences are larger than most people expect.

Why face value lies

A $20 "free gift" and a 10% discount on a $200 order both "cost" $20 on paper. They don't cost you the same.

• Gift: costs you wholesale COGS — maybe $6 for a $20-retail item.
• Percentage discount: costs face value off margin. 10% on $200 is $20, period.
• Free shipping: costs the carrier rate, but only on orders that needed it.

Same headline, different cost. And none of that tells you whether the offer created a sale.

The ranking metric: CPIO

Cost per incremental order (CPIO) = total promo cost ÷ incremental orders vs. control.

Lower is better. The gift's edge shows up here, because its perceived value ($20) far exceeds its cost ($6), while a percentage's perceived value equals its cost by definition.

Worked example (illustrative — plug in your own)

10,000 eligible sessions per arm, $100 AOV, 60% margin. Baseline conversion 2%.

Gift (retail $20, cost $6):

• Treatment accepts 8% → 800; control "buys anyway" 1% → 100. Incremental = 700.
• Promo cost = 800 × $6 = $4,800. → CPIO = $6.86.

Blanket 10% off:

• Treatment converts 2.2% → 220; control 2.0% → 200. Incremental = 20.
• Promo cost = 10% × $100 × 220 = $2,200. → CPIO = $110.

Same broad goal, ~16× the cost per incremental order. That's the whole argument for ranking by CPIO: the "cheaper-looking" discount was the expensive one.

Free shipping lands in between and depends entirely on where it applies: cheap per incremental order when it only bites on below-threshold baskets that wouldn't have converted; expensive when it subsidizes orders already over the line.

Two caveats before you chase CPIO

• CPIO only ranks offers that clear the contribution guardrail. A gift can be cheap per incremental order and still lose money if its incremental contribution doesn't beat the discount it replaced. Always pair CPIO with incremental contribution.
• Acceptor metrics prove nothing. "Gift buyers spent more!" is selection bias — acceptors always do. Measure against a holdout.

How to choose

1. To acquire new customers cheaply: a gift behind a threshold, set 20–30% above AOV, with gift retail worth 5–15% of the threshold. It looks generous and costs wholesale.
2. To move a hesitant returner: a small percentage — gated to hesitation, not blasted storewide.
3. To clear a shipping-cost block: free shipping, only where shipping is the actual reason.

Fund whichever tactic has the lowest CPIO that clears the incremental-contribution guardrail — measured against a holdout, never pre/post.

The takeaway

Stop ranking offers by how generous they look. Rank by cost per incremental order:

A $20 gift can cost $6. A 10% discount costs exactly what it says. Free shipping costs whatever the carrier charges. Until you measure incrementality, you're comparing the wrong numbers.