Custom surveys

NPS, CSAT and CES answer three well-defined questions. A lot of what a team wants to know is none of them: what a visitor came to do, which plan they are comparing, why a form felt long, whether a new page made sense. A Custom survey is built for that, with any mix of score, single choice, multiple choice and open text questions.

It starts from a short example rather than an empty page, a rating, a multiple choice question about why the visitor came and an open question at the end, so you edit towards what you need instead of designing from nothing.

Each score question picks its own scale: nought to ten, one to five stars, or one to seven. A Custom survey can hold several score questions or none at all, which is why its results are read per question rather than rolled up into one headline figure that would mean nothing.

Everything else works exactly as it does for the instruments: the same targeting and triggers, the same frequency rules, the same results by variant when the survey is tied to an experiment. A draft can switch between Custom and an instrument; once a survey has collected answers its type is fixed, because those answers were given on its scales.