Skip to content

UX metrics

Measure NPS, CSAT and UEQ directly in your shop

Three standard instruments with their original scales: NPS from 0 to 10, CSAT from 1 to 5, UEQ with 26 or 8 items. You deliver them as a modal or inline embed and see scores, distributions and comments in the dashboard.

No extra code: NPS, CSAT and UEQ run on the single uxspire snippet that surveys and studies also use.

Active NPS +124w avg 7.5 306 resp · 31 Aug 2026
Active CSAT 71.8%4w avg 3.86 181 resp · 31 Aug 2026
Active UEQ 0.754w score · −3 to +3 71 resp · 31 Aug 2026

3 active instruments · 4-week rollup

All measurements in one project: for each method, the overview shows the value for the last four weeks, the number of responses and the date of the latest response.

Three standard instruments

NPS, CSAT and UEQ have been in use for years. You adopt them with their original scale and the usual calculation, without designing them yourself. Give an example answer and see how it feeds into the calculation.

Net Promoter Score (NPS)

A single question from 0 to 10. The NPS is the share of promoters (9 to 10) minus the share of detractors (0 to 6), on a scale from -100 to +100.

Use case
How likely customers say they are to recommend your shop, asked for example in the customer account or after an order. Over quarters, the score shows how this stated willingness develops.
Limit
The score does not explain causes. Only compare measurements taken at the same moment with the same audience; comments show what lies behind it.

Preview in the shop

How likely are you to recommend us?

How it is calculated

Worked example: the latest 40 responses.

  1. Detractors 0 to 630.0%12
  2. Passives 7 to 830.0%12
  3. Promoters 9 to 1040.0%16

NPS = share of promoters minus share of detractors
40.0% − 30.0% = 10.0

NPS +10

Passives count in the denominator. The NPS ranges from −100 to +100 and is rounded to a whole number.

Customer Satisfaction (CSAT)

A rating from 1 to 5, as a number or an emoji. It measures satisfaction with one specific experience, for example after checkout.

Use case
Satisfaction with a specific experience, such as ordering on the confirmation page or an advice page.
Limit
Satisfaction right after shopping is not satisfaction after delivery. For questions after unpacking, use a survey shared by link.

Preview in the shop

How satisfied are you with your experience?

How it is calculated

Worked example: the latest 40 responses.

  1. 😡 15.0%2
  2. 🙁 27.5%3
  3. 😐 317.5%7
  4. 🙂 437.5%15
  5. 😍 532.5%13

Scale mean 3.85 average of ratings 1 to 5

Top-2 box 70.0% 28 of 40 responses rated 4 or 5

Scale mean and top-2 box describe the same responses in two ways. Always state which value you mean.

User Experience Questionnaire (UEQ)

UEQ with 26 items or UEQ-S with 8 items, both as pairs of opposites. Results range from -3 to +3: six scales for the UEQ, pragmatic and hedonic quality for the UEQ-S.

Use case
How visitors experience using your shop, for example before and after a redesign of the product page or checkout.
Limit
The UEQ measures perceived quality, not conversion. Values are comparable when the variant, page, timing and type of survey stay the same. The UEQ handbook describes how the scales are built and interpreted.
Variant

Preview in the shop · UEQ-S

Please rate the checkout using the following word pairs.

obstructive to supportive
complicated to easy
inefficient to efficient
confusing to clear
boring to exciting
not interesting to interesting
conventional to inventive
usual to leading edge

How it is calculated

Worked example: the latest 20 responses.

8 of 8 pairs still open. The calculation starts once the answer is complete.

Pragmatic quality pairs 1 to 4
+0.98positive
Hedonic quality pairs 5 to 8
+0.63neutral
Overall value
+0.80

Each pair is converted from 1 to 7 into −3 to +3. Means above 0.8 count as positive, below −0.8 as negative. The UEQ handbook does not recommend reading the overall value as a general UX KPI.

With JavaScript enabled you can fill in the instruments and filter the analysis.

Home-made questionnaires cannot be compared

If the question or the scale changes between two measurements, the comparison is void. Standard instruments have fixed questions, fixed scales and a fixed calculation.

Variant fixed once published

You choose the variant before the first publication; after that it stays the same for this measurement so that scores remain comparable. The UEQ word pairs are fixed per language.

Detailed analysis with trend tracking

Each method has its own dashboard: the trend over time, the distribution of response groups, and the comments with date, device type and score group.

329 responses Last 30 days, 2 August to 31 August 2026

Result KPIs

Desktop+16Avg 7.6
Mobile+7Avg 7.3
Tablet+12Avg 7.3

Detailed analysis

Trend analysis

per day

Segment distribution by device

Filtered response distribution by device

  • DesktopDetractors 35, Passives 46, Promoters 57
  • MobileDetractors 46, Passives 52, Promoters 57
  • TabletDetractors 7, Passives 9, Promoters 10
  • UnknownDetractors 3, Passives 3, Promoters 4

DetractorsPassivesPromoters

Voice of the customer

12 of 76

Word cloud

  • 31 Aug 2026, 13:21Rating: 8 · Passives · Mobile

    Nice products, but delivery could be faster.

  • 31 Aug 2026, 11:01Rating: 10 · Promoters · Desktop

    Customer service helped me straight away with a complaint.

  • 30 Aug 2026, 08:50Rating: 9 · Promoters · Desktop

    Great quality, the shelf looks just like in the photos.

  • 30 Aug 2026, 17:44Rating: 10 · Promoters · Desktop

    The chat advice helped me choose the right size.

  • 30 Aug 2026, 19:21Rating: 9 · Promoters · Mobile

    Fast delivery and careful packaging.

  • 30 Aug 2026, 17:20Rating: 5 · Detractors · Mobile

    The delivery came without notice and nobody was home.

  • 29 Aug 2026, 15:51Rating: 5 · Detractors · Desktop

    Delivery took two weeks, one week was announced.

  • 29 Aug 2026, 17:51Rating: 9 · Promoters · Desktop

    Large selection of lamps and clear descriptions.

  • 28 Aug 2026, 09:22Rating: 9 · Promoters · Tablet

    The chat advice helped me choose the right size.

  • 28 Aug 2026, 16:44Rating: 10 · Promoters · Mobile

    Good price for the quality, and the colour is exactly right.

  • 27 Aug 2026, 17:51Rating: 5 · Detractors · Desktop

    The price does not match the quality, the seam is already coming loose.

  • 27 Aug 2026, 17:21Rating: 4 · Detractors · Mobile

    The return was a hassle, I had to print the label myself.

Reading example: In the example, delivery times were added to the product pages on 12 August. Before that the NPS in this slice was +3 (119 responses), afterwards +16 (210 responses). That is a change over time, not proof of a cause: season, promotions, delivery times or different respondents can explain it just as well.

Trend analysis

Filter by time period and device type, and for NPS and CSAT also by response group. That way you compare NPS groups, CSAT ratings and device types. The UEQ follows every filter except the response group.

Comment analysis

Word clouds and comment lists connect open responses with date, device type and score group, from detractors to promoters. Via the response group you get from the score to the individual comments.

Bring existing scores

Historical NPS and CSAT responses from other tools can be imported via CSV, so your time series does not start from zero.

More filters

Narrow the responses by campaign, origin, page, visitor type, country, browser and operating system. Campaign and origin come from the page where the visit began and apply to the whole visit, even if someone answers on a later page. uxspire derives the country from the IP address and stores only the country code. Active filters appear as chips that you remove one by one.

Groups of measurements

You combine several NPS, CSAT or UEQ measurements into a group, for example the NPS on the order confirmation and in the customer account; UEQ and UEQ-S stay separate. The group value counts every response as if it came from one measurement. The group has its own page with its composition, the gap between each measurement and the group value, and the comments from all measurements.

CSV export

You export every measurement as CSV, one row per response. A group goes into one file with all its measurements and a column for the measurement. The file follows the active filters.

Long or short: two versions with different depth

The handbook recommends the short version for situations in which the full questionnaire cannot be used, for example because there is only room for a few questions.

Scroll the table sideways.

AspectUEQUEQ-S
Length26 pairs of opposites8 pairs of opposites
DimensionsAttractiveness, perspicuity, efficiency, dependability, stimulation, noveltyPragmatic and hedonic quality
Suitable forThorough evaluation, for example before and after a relaunchSituations where the full questionnaire is too long; the short version does not measure detailed qualities such as efficiency or novelty
Overall valueThe six scales are interpreted separatelyMean of all eight pairs; according to the handbook, read it as an overall figure only with caution

Structure and interpretation based on the UEQ handbook by Martin Schrepp, version 11 of 12 September 2023, pp. 2, 5 and 11 to 13 (ueq-online.org, accessed on 13 September 2026). In uxspire, the pairs of opposites and instructions are fixed per language.

Settings per measurement

Placement, texts, language and release status are set individually for each measurement.

Precise placement
Deliver the measurement as an overlay, or embed it into a page area using a CSS selector. URL patterns, inline positions and frequency limits decide where and how often it appears.
Texts
The welcome screen, thank-you page and widget texts such as “Submit” can be adjusted. For the UEQ, the instructions and the comment question are editable too; the word pairs stay fixed.
Two languages
NPS, CSAT and UEQ are available in German and English. You set the language per measurement; copy, scale labels and interface elements follow it.
Status
Draft, published, halted or archived. Changes only go live when you publish. Halting stops delivery; the responses collected so far are kept.
Consent first
Embedded measurements only collect answers once your consent management allows it. The CMP integration describes the setup.

What metrics do not replace

You measure orders, returns and contribution margins in your shop and inventory systems. uxspire does not automatically link metrics to revenue or to the segments of an audience study.

The combination becomes useful when you plan it deliberately: an audience study shows which group misses which information. After the change, you watch orders and product questions in your systems and perceived quality with the UEQ-S on the same page. The guide UX research in e-commerce (in German) shows which method answers which question.

Questions

Frequently asked questions about UX metrics

What is the Net Promoter Score (NPS)?

NPS measures willingness to recommend with one question on a scale from 0 to 10. The share of answers from 9 to 10 minus the share from 0 to 6 gives a score from -100 to +100. The score does not explain causes; that is what the comments are for.

What is the Customer Satisfaction Score (CSAT)?

CSAT measures satisfaction with a specific experience on a scale from 1 to 5. It is reported as the average and the share of the two top ratings (top-2 box). Satisfaction right after shopping is not the same as satisfaction after delivery.

What is the User Experience Questionnaire (UEQ)?

The UEQ captures perceived usability through 26 pairs of opposites such as “obstructive” versus “supportive”, and condenses them into six dimensions with values from -3 to +3. The short version UEQ-S uses 8 pairs and shows pragmatic and hedonic quality.

How are the results analysed?

As trends by period (7, 30 or 90 days or all time); for NPS and CSAT also by device and response group. Comments can be read with a word cloud of frequent terms. Historical NPS and CSAT scores can be imported via CSV.

Plan your first measurement.

Three standard instruments, one view. As a modal or inline embed, in German and English.


Standard scales Modal or inline German and English EU hosting