First-Order Discounts: Do They Buy Loyal Customers or Bargain Hunters?

First-Order Discounts: Do They Buy Loyal Customers or Bargain Hunters?

Offering a first-order discount is near-universal advice. Popup apps sell it, agencies recommend it, every competitor seems to do it.

Almost nobody checks whether it works — because that requires comparing the customers it brings in against the customers it would have brought in without it.

The question isn't "does it convert?"

Of course a discount converts more first-time visitors. That's not the question.

The questions are:

1. Would those customers have bought anyway? (incrementality)
2. Do discount-acquired customers come back? (cohort LTV)
3. Does the discount pay for itself on the first order, or only on repeats? (CAC payback)

First-order payback math

Let:

• B = first-order basket value
• m = contribution margin rate
• S = shipping cost you bear
• CAC = fully loaded acquisition cost

First-order contribution:

π₁ = B·(m − d) − S − CAC

The maximum first-order discount that breaks even:

d_max = m − (S + CAC) / B

Worked example. Basket $80, margin 60%, shipping $7, CAC $30.

d_max = 0.60 − (7 + 30)/80 = 0.60 − 0.4625 = 13.75%

• At 10% off: first order contributes $3 → pays back on its own.
• At 13.75% off: exactly break-even.
• At 20% off: first order loses $5 → it's an investment requiring repeat orders.

So the standard "20% off first order" is losing $5 before a single repeat. It's only profitable if those customers return — and return at a margin that isn't also discounted.

Now add incrementality

The payback math above assumes the discount acquired the customer. But many would have bought anyway.

Say only 40% of the discount-acquired customers are truly incremental. Your effective CAC becomes:

CAC_effective = $30 / 0.40 = $75

New break-even:

d_max = 0.60 − (7 + 75)/80 = 0.60 − 1.025 = negative

Meaning: at a realistic incrementality rate, no first-order discount pays back unless the customers it brings are worth far more in repeat value than the ones you'd have acquired anyway.

The cohort reality

This is the part that should worry you. Run a holdout — give no welcome offer to 5% of visitors — and track both cohorts over 12 months. What stores commonly find:

• Discount-acquired customers repeat less. They came for the deal; the next deal is a competitor's.
• They wait for codes. Having learned that "first order = 10% off," they extend the pattern: wait for a sale, search for a code, expect a discount every time.
• Their LTV is lower, not higher, in many categories — because the discount didn't buy loyalty, it bought a transaction.

Combine that with a compressed first-order margin and the "acquisition" can be worth less than a full-price first order.

How to actually measure it

1. Holdout. Withhold the welcome offer from a small random slice. Compare incremental signups and first orders.
2. Cohort LTV. Track 12-month contribution of discount-acquired vs. full-price-acquired vs. holdout cohorts. Include the discount on all their repeat orders.
3. Blended CAC vs. incremental CAC. Report both. Only the incremental number tells you what the discount bought.

The fix

If the cohort is weak, the answer isn't to remove the welcome offer — it's to change what it offers:

• Give value, not just price. Free shipping, a welcome gift, or a service (returns, support) can convert without training code-waiting.
• Gate the discount. Trigger it at hesitation, not automatically for everyone. Then you're not teaching a habit — you're solving a moment.
• Reserve price cuts for the recalcitrant. Some shoppers need a discount to try you. Most don't. Stop treating "new" as synonymous with "needs a discount."

The takeaway

A first-order discount is an acquisition investment. Investments need a return — and the return is repeat contribution, in a cohort you can actually compare to a holdout.

If the discount-acquired cohort repeats less and waits for codes, you didn't buy loyal customers. You bought a habit — of waiting for a discount.