One-dimensional Quality

Teams used to allocate effort on the assumption that improving something makes people happier. Results didn't cooperate. Cutting document load time in half moved the satisfaction survey; tripling the number of options on the settings screen moved nothing at all. Same effort spent on both, and only one of the two responded at all.

With no name to separate the two, both sat on the same line in the same meeting. "Surely it gets better if we make it better" was impossible to argue against, so the order got decided by whoever pushed hardest. The Kano model split features into five types, and one-dimensional quality is the name given to the only one where investment and satisfaction run in a straight line.

Put more in and satisfaction goes up by roughly that much; take some out and it falls by roughly that much. It is the only type that moves in both directions, which means what belongs here is anything with a scale — seconds to open a document, the accuracy of search results, the storage a free plan includes.

Two questions decide it. Ask how someone would feel if the feature were present and they say they'd like it; ask how they'd feel without it and they say they'd dislike it. Answers of "that's just expected" when present point to must-be quality, and answers of "no strong feeling" when absent point to attractive quality.

The straight line comes with a cost: a small gain buys a small amount of satisfaction. Taking document load time from 3 seconds to 2.8 raises the number on paper and nobody notices. The threshold where people actually feel a change usually means halving a figure or doubling it, so a quarter spent entirely here can end with a better number and nothing that anyone can see.

"This is one-dimensional, so let's set a target number before we start."

The attribute has a scale, so decide how far it should move first.

"Nobody feels 3 seconds becoming 2.8."

The relationship is linear, but this step is too small to cross into perception.

"Competitors are at 2 seconds and we're at 5. This is where we get compared."

Numbers can be lined up side by side in this type, so the gap shows plainly.

Pagedeck classifies search accuracy as this type and sets a target figure each quarter. Last quarter Mia wrote down moving the share of searches that surface the right document on the first screen from 62 percent to 81 percent, and Theo scoped the work against that number. The reason it wasn't 70 percent was a calculation done first: at 70, users wouldn't have noticed the work at all.