Kiosk vs QR vs email: 2026 customer-feedback comparison
27 octobre 2026
Collecting a customer's opinion in 2026 is no longer the same job as in 2016. Satisfaction emails are tiring, QR codes are everywhere but convert poorly, and touch kiosks are gaining ground in venues where physical presence is the main asset. Each method has a sweet spot, and picking the right one for your context saves months of pointless collection.
This article compares the three dominant collection methods across seven concrete criteria: response rate, cost per response, data quality, immediacy, technical reliability, compliance, and sampling bias. The goal is to give you a decision framework before you invest, not to push one format in particular.
The three methods in thirty seconds
Quick recap, because the terminology often blurs in specs:
- The touch kiosk is a screen fixed on site, standalone, that asks one to three questions to the customer right after their experience. No phone, no email, no app to install.
- The QR code is a printed square on a receipt, a table card, or a window, redirecting to a web form on the customer's smartphone.
- The email is a message sent a few hours or days after the transaction, carrying a link to a web form.
All three aim at the same thing, an actionable opinion, but tap three very different attention models.
Response rate: what actually works
Rough ranges, calibrated on our field data from the last two years:
| Method | Median response rate | Observed range |
|---|---|---|
| Touch kiosk (exit) | 35% | 25 to 55% depending on spot |
| QR code on receipt | 2 to 5% | 0.5 to 8% depending on visual and incentive |
| QR code on table | 4 to 8% | 2 to 15% depending on context |
| Email (D+1 to D+3) | 10 to 15% | 5 to 25% depending on base |
| Email with incentive | 15 to 25% | 10 to 35% depending on voucher |
The lesson: the kiosk wins the raw-rate race hands down, but only against customers physically present. Email remains the only method that reaches remote customers and e-commerce. The QR code sits in an uncomfortable middle, often too much friction for its audience.
Cost per response: the real trade-off
The gross cost of a method is misleading. What matters is the cost per usable response.
- Touch kiosk. Upfront investment of 1,500 to 3,500 euros, software at 30 to 90 euros per month. In a shop generating 1,500 responses per month, cost per response drops to 10 to 30 cents after year one.
- QR code. Technical cost near zero (a QR generator, a web form). Real cost: the human resource that has to prompt, embody the QR in the relationship, and process returns. Usually underestimated.
- Email. Cost of the sender platform (20 to 200 euros per month depending on volume), plus base maintenance cost. On a 10,000 customer list, a monthly send costs 50 to 100 euros, for 1,000 to 1,500 responses, or 5 to 10 cents per response.
Raw comparison: email is cheapest per response at scale, kiosk becomes competitive as soon as the physical customer base is sizeable, and QR code is rarely the most cost-efficient option if you count the human activation time.
Immediacy: catching the right moment
This is where the three methods diverge the most.
The kiosk catches the customer within the second following their experience. Emotion is sharp, detail is fresh, the answer reflects what just happened. This is precious for spotting a weak signal (a rude cashier, an abnormal wait), because the fix can land that same day.
The QR code sits in between. If the customer scans from their table or receipt immediately, you are close to real time. If they scan the next day, the signal dilutes in their memory.
The email is lagged. An email sent at D+1 captures a stabilised feeling, sometimes shifted by post-visit rumination or a conversation with a relative. Useful for topics that need reflection (a purchase decision, medical advice), less for operational topics (reception, cleanliness, waiting time).
Data quality: personal smartphone or fixed tablet
Response quality depends on the medium as much as on the question.
On a fixed touch kiosk, the response is brief, the customer has a few seconds and will not type a paragraph. You get nice quantitative data (scores, multiple choice), but few long verbatims.
On a smartphone, via QR code or email, the customer can theoretically write a longer response. In practice, 85% of answers fit in under 20 words. The smartphone is a reading medium, not a writing one.
For rich qualitative verbatims, only one format works: a targeted follow-up email to respondents who already answered positively and agreed to be recontacted. It is both slow and expensive, but it is the only reliable way to get usable material for a proper study.
Technical reliability
A criterion often forgotten until the first incident:
- Kiosk. Depends on internet and power. On an outage, serious software queues answers locally and ships them on reconnection. Less serious software drops answers silently.
- QR code. Depends on the customer's smartphone, their network, the native scanner, the right URL. Each link adds friction. A badly printed QR code or one in a glare becomes unusable.
- Email. Depends on deliverability, meaning the share that lands in the primary inbox and not in spam. A badly configured email in SPF or DKIM can lose 30% of sends silently.
All three have silent failure modes, but the kiosk is usually the easiest to monitor, because you can look at it.
GDPR, consent, bias
Each method has its own list of legal and statistical traps.
- Kiosk. No email required for the main question, which simplifies GDPR. If you ask for email for a receipt or follow-up, explicit opt-in via checkbox. Watch for self-selection bias: customers in a hurry or in a group do not stop, and their profiles are systematically under-represented.
- QR code. Few personal data if you stay anonymous. Strong device and generational bias: an older or low-bandwidth audience scans much less. For a regional museum or a town hall, this is a real problem.
- Email. Requires a customer base with GDPR consent. Strong representation bias: your addresses usually come from those who booked online, bought on site with a loyalty card, or subscribed. Occasional visitors are missed.
In practice, no single method gives a truly representative sample. That is why the best strategies combine at least two channels.
Which method for which context
A quick decision table:
| Context | First method | Useful second channel |
|---|---|---|
| High-footfall physical retail | Kiosk at exit | Monthly email to loyalty cards |
| Restaurant | QR code on table | Email after online booking |
| Public service, administration | Kiosk at reception | QR code in waiting room |
| Pure e-commerce | Email at D+3 | Quarterly open survey |
| Hotel or residence | Post-stay email | QR code in the room |
| One-off event | Mobile kiosk | QR code on badge |
| Clinic, medical practice | Kiosk in waiting room | Email follow-up to non-respondents |
The rule that works: start with the channel where the customer is most captive, and layer a secondary channel that catches the rest.
Winning combinations
The best setups we see at our clients weave several methods into a single coherent journey.
- Retail combo. Kiosk at exit for the real-time signal, monthly email to the loyalty base, QR code at the till for customers in too much of a hurry to stop at the kiosk. A single back-office analytics tool aggregates the three sources.
- Administration combo. Kiosk at reception to measure waiting, QR code in the waiting room for users who prefer silence, monthly email to recurring users (document renewals, requests). Separates one-off irritants from structural issues.
- Hospitality combo. QR code in the room for in-stay irritants, post-stay email for the overall view, kiosk in the restaurant reception. Each channel answers a specific moment.
In all cases, the common success factor is the same: an owner who reads all three sources together every week and discusses the gaps with the team.
Measuring ROI
Three indicators to set before any deployment, to avoid projects that drift without anyone knowing whether they work:
- Response rate per channel. Measured as respondents divided by eligible passages or sends.
- Cost per usable response. Full device cost divided by annual response count, duplicates stripped out.
- Action rate. Number of concrete actions taken based on the feedback, divided by the number of alerts identified. This is the most telling indicator: it reveals whether the data serves anything.
A device that has been running for six months without any action following from it is a device to reconsider, whatever the method.
Going further
If you are scoping a concrete project, our customer satisfaction kiosk product page details the available kiosk models, and our article touch screen survey terminal guide covers the kiosk format specifically. For the software side orchestrating the three channels, see customer satisfaction application.
Need a scoping call on your project? A 30-minute conversation is enough to identify the most relevant combination for your context and budget the setup.