Updated
Customers care about the date their order arrives. The carrier and the service that get it there are your problem, not theirs.
It starts at checkout, where the customer expects to be told when the order will arrive and what it costs.
Give them a date they can plan around and the service behind it stops mattering to them.
Most checkouts do the opposite, listing carrier services rather than delivery outcomes.
A checkout offering “USPS Priority Mail”, “FedEx 2-Day”, “FedEx Ground”, “UPS 2nd Day Air”, “UPS Ground” and “USPS Ground Advantage” is trying to be helpful by showing everything. What it actually does is hand the customer a decision they are not equipped to make. Three problems follow.
Collapse the list into two or three categories — Economy, Expedited, Overnight — that a customer can choose between without knowing anything about carriers.
These options should be accompanied by a clear expected delivery day.
Display the arrival date next to each shipping option at checkout.
Give a date, not a duration. “Arrives Tuesday 8 September” rather than “3 days”, because a duration hands the customer a calculation involving your cutoff, your processing time and the weekend, and they will get it wrong. What to offer at checkout and what to charge for it goes through the format in detail.
As a rule of thumb from the operations I have worked in, economy lands in five to seven days, expedited in two to three, and overnight the next day. Those windows leave enough room to choose a service inside them.
Selling a date rather than a service leaves you free to buy whichever service hits that date cheapest.
A worked example. You ship from New Jersey, the customer is in New York City, and at checkout they paid for overnight.
On that lane, UPS Ground arrives the same day as UPS Next Day Air.
The difference is that Next Day Air would cost you $50+ and Ground only costs $12.
The customer gets the package on the day they were promised either way. Had they picked Next Day Air by name at checkout, you would have been committed to it.
Done consistently, that margin is yours to keep, or to spend on offering cheaper expedited shipping at checkout.
If nobody has time to make that comparison order by order, it has to be automated. The useful part is that delivery times are unaffected — you are buying the same arrival date on a cheaper service.
Across the operations we work with, rate and packing optimisation alone runs roughly $1 to $3 an order under about 6,000 orders a month, and $3 to $5 above it. Operations with real complexity find most of the value elsewhere: the manual decisions, the exception queue, and the consistency of getting the same answer regardless of who is on the floor. What those figures actually cover. If you want the number for your own order mix, the self assessment works it out from two ShipStation exports.
A checkout that answers when the order arrives, and what it costs, is easier to buy from than one that asks the customer to interpret a carrier service.
Knowing how a carrier prices a parcel is useful. The harder question is which parcel to make — because the packing decision sets the dimensional weight and the cubic tier, and the cheapest box changes with the contents, the destination and the rate card.
There are eight situations where that decision has grown past what a rule can express. Most operations are in two or three of them.