Shipping you can see through.

It is the share of parcels we can confirm were delivered, out of the parcels that finished their journey, measured only on orders that shipped long enough ago to have arrived. So a 99% line means about 99 in every 100 of those parcels reached the customer, confirmed by the carrier.

Those are parcels we cannot yet confirm as delivered. Almost always the parcel did arrive but the carrier stopped scanning near the very end, which is common on some last-mile networks, so we choose not to count it as delivered until we have proof. It is a gap in the paperwork, not a lost parcel. That is why we show it separately instead of hiding it inside the arrived rate.

Very few. The "did not arrive" number is the real one: genuine returns or failed deliveries, usually under 1%. Everything else either arrived or is still being confirmed. We would rather under-claim than show a number we cannot back.

No. When the carrier tells us a parcel came back because the address was incomplete, the house number did not exist, or nobody could be reached, we leave that parcel out of the line's record entirely. It is not a fair way to judge a courier, and on our biggest route it is over a third of all failures. What we do NOT do is give a carrier the benefit of the doubt: if a parcel failed and no reason was given, it still counts against the line. Otherwise a carrier that tells us nothing would look perfect.

A parcel that shipped last week has not had time to arrive yet, so counting it would unfairly drag the rate down. We wait until every parcel in the window has had enough time to be delivered, then measure. It is the only honest way to score a delivery rate.

Because a lifetime average would tell you about carriers we no longer use and lanes that no longer run the way they did. Carrier performance moves, routes get rerouted, and a line that was excellent two years ago can be mediocre today. Holding the window to 180 days means what you are reading describes how these lines are performing now, which is the only version worth quoting to you. The shipment count at the top is that window, not our total volume.

It blends three things, weighted by what matters most. Reliability, the arrival rate, counts the most. Then speed, the typical transit time. Then how often a parcel arrives within our target window. A fast and reliable line scores high; a reliable but slow line scores a bit lower. That is why a line can be very reliable yet rank below a slightly faster one.

Probably both of you, on different products. A carrier is not one speed. YunExpress alone sells nineteen products into the United States, and the cheapest and the fastest are quoted five to fifteen working days apart, with the fast ones costing thirty to seventy percent more. Our headline figure is the line that carries almost all our parcels, which is an economy line. Each carrier now lists its own lines underneath with a tag showing whether it is an economy or an express product, so you can compare like with like. If speed matters more than cost to you, ask us and we will quote the faster line.

The typical number of days from when a parcel ships to when it is delivered, taken from the real carrier delivery scan. These are calendar days, counting weekends and holidays, not business days, so the number matches the real wait your customer experiences. Median means half of parcels arrive faster and half slower, so one unusually slow parcel does not skew the number.

Because we start the clock earlier. A carrier quotes from the moment they collect the parcel. We start from the moment the label is created, which is typically a day or two before that, and we count every day until it is delivered. Both numbers are on each line above, so you can compare like for like: the carrier portion is the one that matches a carrier quote, and the full figure is what your customer actually waits.

Usually because it is averaging orders that are still in the air. If you take last week of orders and average the ones that have already arrived, you are averaging only the fast ones, because the slow ones have not landed yet and so cannot be counted. Measured on our own data, that method reports about 7 days for a route whose real figure is 10. We only measure parcels that shipped long enough ago for every one of them to have arrived, which is why our number is higher and why it does not move as the week goes on.

Yes. They are computed live from real carrier data, not marketing, and nothing is cherry-picked. You can verify it yourself: the demo above pulls real shipments at random and shows their real tracking number, so you can look them up.

Because that is where we fulfill from, and the left side is the full, unedited journey we track internally. Your customers never see that. They see the clean, on-brand page on the right, where the origin and carrier are masked.

It means we do not yet have enough shipments or detailed scans on that route to state a precise number, so we flag it instead of showing something falsely exact. It fills in as volume grows.