Three numbers for the cabinet package, and they are not for the same job. The screen's work is not arranging the totals, it is producing the fact they hide: level the one difference that carries a price and two of the three land $600 apart.
All to Dave, over two weeks. Nothing is settled and nothing is recommended.
Hiding the three numbers until somebody has read a scope table would be a product deciding it knows better than the person whose money it is. They are the first thing anybody wants and they are at the top. What follows them is the correction, in the same breath, which is the only honest arrangement of a number that is true and misleading at once.
Same rule the option priced past its allowance follows on 13. An allowance is a plan, quotes are the world, and a product that treated the plan as a limit would be arguing with reality on behalf of a spreadsheet written a year ago.
The only emphasis on the screen marks where the three quotes stop agreeing. It never marks one as better, because plywood drawer boxes is better construction and seven weeks is a better date and the product has no basis for ranking one against the other on somebody else's build. Marking the difference is help; marking a winner is an opinion the product has not earned.
Three money columns and a label column do not fit on a phone, and the design system's answer to that is to drop a column rather than shrink the cells. So the label takes its own line above the three, which keeps every figure at full size and keeps the three columns readable as columns. On a wide screen the label rejoins them as a fourth.
Listing what matched matters as much as listing what did not: without it the reader cannot tell whether the product examined finish and hardware and found them equal, or never looked. An omission and an agreement look identical unless one of them is drawn.
“If we add crown to Reed's, does it still come in under the second quote, and does it move the ten weeks?”
His answer lands on the cabinet package, not in a message list. A year from now the question is why these cabinets, and the answer needs to be next to them.
The screen deliberately has no choose this one on it. Settling a cabinet package is a conversation with the person building the house, and a product that offered a one-tap winner would be inviting somebody to commit twenty-one thousand dollars on the strength of a table it built out of three PDFs.
Not a generic prompt. It is derived from the one line where the gap between two quotes is larger than the gap between their totals, which is the question a person would arrive at after ten minutes with the table. Offering it is the coach doing arithmetic rather than the coach having opinions.
It lands in Dave's register beside a sub confirming a date, it carries a time-to-decision, and his reply files onto the decision rather than into a thread. Quotes are not a separate correspondence system, and the moment they become one is the moment the record stops holding why a thing was chosen.