What Carrier Scorecards Miss About the Delivery Promise

Carrier scorecards fail for a boring reason: they measure the carrier, not the promise made to the shopper. The two sound like the same thing. They are not, and the gap between them is where most e-commerce delivery problems hide.

What a scorecard usually measures

Most operations teams grade carriers on transit time and first-attempt success. Both are real, both are lagging, and neither tells you whether the date shown at checkout was honored. A carrier can hit its service-level target every week while your delivery promise misses by two days, because the promise was never built from that carrier's actual performance on that lane.

The typical scorecard also averages across everything a carrier touches. A network-wide on-time figure blends dense metro lanes with remote rural ones, light parcels with bulky ones, and quiet weeks with peak. The average looks stable, so nobody acts on it. Meanwhile, a handful of specific lanes keep breaking the date your shoppers were shown, and the scorecard has no row for them.

There is a timing problem too. Scorecards are usually reviewed monthly or quarterly, long after the orders in question were delivered, refunded or complained about. By the time a lane shows up as a problem, the customers who experienced it have already decided whether to buy from you again.

Why the promise is the right unit

The shopper never sees your carrier contract. They see one thing: the delivery date at checkout. That date is a commitment your brand made, not the carrier, and it is the date the shopper uses to judge whether the order went well.

When you measure the promise instead of the carrier, three questions become answerable that a scorecard cannot answer:

• Was the date we showed realistic for this lane, this service level and this time of week?

• When we missed it, was the cause the carrier, our own warehouse cut-off, or a promise that was padded or optimistic from the start?

• Which carrier would have kept the promise on this lane if we had routed the order differently?

Each of those questions points to a decision you can make this week. A service-level average points to a meeting.

Measure the promise instead

Start from the date the shopper was shown. Compare it to the date the parcel arrived. Split the gap by lane, by service level, and by the day of week the order was placed. Carriers that look identical on a service-level report separate immediately once you grade them against the promise.

Then separate the gap into its parts. Some of it belongs to the carrier: time in transit and failed delivery attempts. Some of it belongs to you: orders that missed the warehouse cut-off, labels created late, parcels handed over after the scheduled pickup. And some of it belongs to the promise itself: a checkout date set from a generic transit table rather than from what actually happens on that route. Without this split, every miss gets blamed on the carrier, and the fix that would actually work never gets made.

Finally, look at both directions of the miss. Late deliveries get the attention, but consistently early deliveries are also a signal. They usually mean the checkout date is padded with safety days, which costs you conversion at the moment the shopper decides whether to buy.

What changes operationally

Three things follow. Routing rules stop being static. The checkout date stops being a guess padded with safety days. And the weekly carrier conversation moves from "your service level slipped" to "your performance on this lane broke a promise we made to a customer," which is a conversation that actually changes behavior.

Routing becomes a lane-by-lane decision. Instead of assigning one carrier per region for a year, you send each order to the carrier most likely to keep the date on that specific route, and you can see when that choice should change.

The checkout date becomes something you can defend. When the promise is built from observed performance on the lane, you can show a tighter date where the data supports it and a more cautious one where it does not, rather than one padded date everywhere.

Carrier reviews get shorter and more useful. A list of the lanes where a carrier broke your promises, with the share of misses that were theirs rather than yours, is a far stronger basis for a conversation than a single on-time percentage both sides already expected.

Where this breaks

It breaks when promise data and carrier data live in different systems, which is the normal case. The checkout date sits in the storefront or order management system, while delivery events sit in carrier portals and tracking feeds, each with its own status codes and time zones. Joining them after the fact is possible but slow, and slow feedback is why most scorecards stay lagging indicators nobody acts on.

The practical fix is to store the promised date with the order at the moment it is shown, and to normalize carrier tracking events into one consistent set of milestones as they arrive. Once both sit side by side, measuring the promise is a daily report rather than a quarterly project.

Where to start

You do not need to rebuild your carrier program to begin. Pick your highest-volume lanes, record the date shown at checkout for every order on them, and compare it with the actual delivery date for a few weeks. Split each miss into carrier, warehouse and promise causes. The first report will usually show that the lanes you worried about are not the ones breaking the most promises, and that is where the useful work starts.

Frequently Asked Questions

What is a delivery promise in e-commerce?

A delivery promise is the delivery date or window a shopper sees before they buy, usually at checkout. It is a commitment made by the retailer, not the carrier, and it is the standard the shopper uses to judge whether the order arrived on time.

Why don't carrier scorecards show delivery promise performance?

Most carrier scorecards measure the carrier against its own service level, averaged across all lanes. They do not compare the actual delivery date with the date shown to the shopper, so a carrier can meet its targets while the retailer's promise is still missed on specific routes.

How do you measure delivery promise accuracy?

Record the promised date with each order when it is shown at checkout, then compare it with the actual delivery date. Split the results by lane, service level and order day, and separate each miss into carrier, warehouse and promise-setting causes.

Are early deliveries a problem?

Consistently early deliveries usually mean the checkout date is padded with safety days. A padded promise can make the offer look slower than it really is, which can weigh on conversion at checkout.

Where should a retailer start?

Start with the highest-volume lanes. Capture the promised date for every order on them, compare it with delivery outcomes for a few weeks, and use the misses to adjust routing rules and the dates shown at checkout.