Digital feedback surveys collect opinion at the point of experience: the smiley-face terminal by the exit, a question on a kiosk, or a scanned link at the table, rather than by email three days later.
Timing is the whole advantage. Response rates at the moment of leaving are far higher than post-visit email, and the memory is accurate rather than reconstructed. A single question answered by hundreds beats a detailed survey answered by four.
The caveat is severe response bias, and it gets ignored because the volume looks reassuring. People who press a button on the way out are disproportionately delighted or angry; the large indifferent middle walks past. Treat the results as a trend line for the same location over time, not as a satisfaction score, and never compare two sites whose terminals are placed differently.
Keep it to one question, or possibly two. Every additional question roughly halves completion, and a five-question survey at an exit collects data almost exclusively from people with a complaint.