Where to Set Your Signature-Required Threshold: A Break-Even Model for High-Value Orders

Requiring a signature on every order is expensive and hurts delivery. Requiring none leaves your biggest orders exposed to item-not-received disputes. The right answer is a threshold, and you can calculate it from three inputs.

What a signature changes

When tracking says delivered and the customer says nothing arrived, a signature is the strongest evidence you can show. Community discussion among merchants is consistent on this: without a signature on a high-value order, a delivered scan alone often fails to win the dispute.

Marketplaces already encode this. eBay, for example, requires signature confirmation on orders over $750 for seller protection. That is a marketplace policy, not a rule for your store, but it shows where large platforms draw the line.

The break-even formula

A signature is worth requiring when the expected loss it prevents is larger than what it costs. Written out:

Order value x dispute rate x (loss rate without signature minus loss rate with signature) > signature fee + delivery friction cost

Solve for order value and you get your threshold:

Threshold = (signature fee + friction cost) / (dispute rate x loss-rate improvement)

The three inputs, explained

• Dispute rate: the share of delivered orders that end in an item-not-received dispute. Pull it from your last 12 months of payment disputes and delivery data.
• Loss-rate improvement: how much more often you lose a dispute without a signature than with one. Use your own history. If you have no signed orders yet, start with a conservative assumption and revise.
• Friction cost: the price of missed deliveries, redelivery attempts, customers who abandon the checkout because they must be home, and extra support contacts.

A worked example with assumed numbers

These numbers are an illustration, not benchmarks. Assume the signature fee is $5 and friction costs $3 per order, so the cost is $8. Assume 1% of orders become disputes, and signature changes the loss rate from 70% to 20%, an improvement of 50 percentage points.

Threshold = $8 / (0.01 x 0.50) = $1,600.

Now assume a higher-risk category with a 2% dispute rate. The threshold falls to $800. For a store with a 0.25% dispute rate, it rises to $6,400. The point is that your threshold is set by your own dispute history, not by a number copied from a marketplace.

Adjust for the things the formula ignores

• Brand and category: items that are easy to resell, such as electronics, jewelry and watches, attract fraud and justify a lower threshold.
• New versus repeat customers: a first-time buyer with a mismatched billing and shipping address is riskier than a repeat customer shipping home.
• Shipping destination: apartment buildings and shared mailrooms create more delivery disputes than single-family homes.

Use tiers instead of one number

One threshold is simple, but tiers are more accurate. A common structure is no signature below your lower line, signature required above your upper line, and a review step in between that looks at customer history and address type.

Set a reminder to recalculate every quarter. Dispute rates change with seasons, and a threshold that worked in July may be wrong in December.

Pair the signature with a resolution path

A signature stops disputes on the largest orders but does nothing for the rest. For orders below the line, the cost of a lost or stolen package should be handled before it becomes a chargeback.

Explore the ShipAid Shipping Guarantee to give customers a resolution inside your own branded experience, so the orders under your threshold do not turn into disputes with the card network.