Comparing carrier rates at the moment of despatch — and why the cheapest quote is frequently the wrong answer.
Rate shopping requests a price from each available carrier and service for a given shipment, then selects one according to your rules. The naive version picks the cheapest. The useful version weighs cost against committed delivery date, customer promise, carrier reliability on that lane and any contractual volume commitment.
Rate shopping reduced to 'pick the cheapest' tends to erode service quietly. A two-day saving of a small amount per parcel looks sensible in aggregate and produces late deliveries against a promise the customer was given at checkout. The business rules are the point — the rate comparison is just the input to them.
Weight, dimensions, destination, declared value, required delivery date and any special handling.
Live from each carrier API, or from negotiated rate tables held locally where API latency matters at volume.
Service level must satisfy the promise; preferred carriers on certain lanes; volume commitments; exclusions for restricted goods.
The chosen service is booked and the label produced in the same transaction.
Storing the quotes that were not chosen is what makes later carrier negotiation evidence-based.
A 4 kg parcel measuring 50 by 40 by 35 centimetres goes to a tier-two city. Carrier A quotes 240 on actual weight. Carrier B quotes 265. Carrier C quotes 290 with a next-day commitment.
Dimensional weight changes the picture before anything else. At a common divisor the parcel's volumetric weight is about 14 kg, well above its actual 4 kg, so all three price on 14 kg and the gaps widen. Then the remote-area surcharge applies to A and B but not to C, which has a depot in that city. Then serviceability: A's measured on-time rate to that zone is 82 percent against C's 97 percent, and each failed delivery costs roughly 400 in return freight and handling.
Expected cost rather than quoted cost reverses the ranking. A's 18 percent failure rate adds about 72 of expected return cost, which more than erases its apparent advantage. The useful output of rate shopping is not the cheapest quote — it is the lowest expected total, which requires dimensions, surcharges and measured performance to be in the calculation.
| Term | What it means |
|---|---|
| Rate card | The negotiated price table for a carrier, by service, zone and weight band. |
| Zone | The carrier's geographic banding, which drives price far more than distance does. |
| Surcharge | Additional charges beyond the base rate — fuel, remote area, residential, oversized. |
| Service level | The committed transit time, which is what makes the cheapest option sometimes unusable. |
| Selection rule | The logic choosing a carrier, which should encode more than price alone. |
Missing a delivery promise costs more than the shipping difference, in refunds and in repeat business.
Locally held rates drift from contract. Refresh them on a schedule or rate live.
Live rating in the customer's checkout adds delay. Cache or estimate at checkout and rate properly at despatch.
Without the rejected quotes you have no evidence for the next carrier negotiation.
| Metric | What it tells you | What distorts it |
|---|---|---|
| Realised vs quoted cost | Whether your rate cards and dimensional rules reflect reality. | |
| Selection distribution | Which carriers actually win. Heavy concentration on one suggests the rules are not discriminating. | |
| Expected vs actual failure cost | Whether performance weighting is calibrated or guessed. | |
| Dimension data completeness | Items with no dimensions, each of which forces an approximate quote. |
We build and support the systems this page describes. Start with a free working proof of concept — no cost, no commitment.