How to measure customer satisfaction in a library

Library visitors are mostly anonymous — many don’t have an account, and those who do rarely answer an emailed survey days later. An exit kiosk catches people at the moment the visit is fresh, without asking for a name or an email. It also tells you which branch the answer is about, which an online survey usually can’t.

The trade-off is attention: someone walking out with an armful of books will give you five to fifteen seconds, not five minutes. Everything below is designed around that.

Start with the one question that summarises the whole visit, then go from the most actionable statement to the least. Every statement after the first is answered on a 0–10 scale, from 0 “strongly disagree” to 10 “strongly agree”.

Why statements instead of questions? “I find it easy to find my way around” can be answered in a single tap on the same 0–10 scale as every other row, so visitors never have to re-read instructions. Keep every statement positive and about one thing only.

On a Sensa kiosk, each answer is saved the moment it’s tapped — not when the visitor reaches the end. If someone answers three questions and walks away, those three answers are kept. After about a minute without a tap, the kiosk quietly resets to the first question, ready for the next visitor.

That has two consequences:

• Question 1 gets the most answers, and each later question gets a few fewer. That’s expected and fine — it’s why the most important question goes first.
• There’s no “abandoned survey” to throw away. A partial answer is still a real answer. You don’t need a reward or a progress bar to get value from people in a hurry.

Compare averages between questions, not answer counts. A question with 250 answers and one with 400 are both solid samples.

Here’s an illustrative month at one branch, rounded and modelled on the patterns we saw in a real library pilot:

Three things jump out:

• Staff are the strength. 9.1 is a number to share with the team.
• Orientation is the weak spot. At 6.2 it’s the lowest score by a clear margin, even though most visitors still find what they came for. People are getting there, but it’s harder than it should be.
• The headline NPS is healthy, so nothing is on fire — which makes this the right moment to fix wayfinding before it costs loyalty.

1. Find the lowest average with at least 30 answers behind it. That’s your first candidate.
2. Check it against the NPS. Export the responses and compare: if visitors who gave a low recommendation score also scored that statement low, fixing it will move loyalty, not just one number.
3. Compare branches. If one branch scores 8 on orientation and another 5, visit the good one and copy what it does.
4. Make one change — clearer signage at the entrance, a floor map, colour-coded sections, staff stationed near the entrance at peak hours.
5. Leave the survey running unchanged and watch that one average over the next four to six weeks. Don’t reword the question in the meantime, or you can’t compare before and after.

1. Write your survey with the seven questions above, NPS first. Keep the wording identical across branches so the results are comparable.
2. Create one location per branch, so every answer is tied to the place it came from.
3. Put a tablet at each exit, at chest height, next to where people walk out rather than behind a desk. Any iPad or Android tablet works — see our guide to locking an iPad in kiosk mode.
4. Add a small sign: “How was your visit? One tap per question, anonymous.”
5. Wait for about 30 answers per question before reading anything into averages; at a busy branch that’s a few days, at a small one a few weeks.
6. Review monthly — one look at the lowest score, one decision, one change.