UX Design · Lesson 13 / 13
Measuring UX
How to measure user experience with behavioural and attitudinal metrics — task success, SUS, CSAT, NPS — plus Google's HEART framework, funnels, A/B testing and how to choose metrics that matter.
5 min read · 5 quiz questions
Why measure UX?
Design decisions are easier to defend, and improvements easier to prove, when you can measure them. Metrics tell you whether a redesign actually helped, where people struggle at scale, and how your product compares over time.
Measurement also connects UX to business outcomes. Saying "the new checkout feels better" is an opinion; saying "task success went from 72% to 91% and completed orders rose 8%" is evidence.
Behavioural metrics
Behavioural metrics measure what people actually do. You can collect them in usability tests or through product analytics.
| Metric | Definition | Example |
|---|---|---|
| Task success rate | % of users who complete a task correctly | 85% of testers found the return policy |
| Time on task | How long a task takes | Average checkout time dropped from 3m 10s to 1m 50s |
| Error rate | How often users make mistakes | 1 in 4 users entered the date in the wrong format |
| Conversion rate | % who complete a goal (sign up, buy) | 3.2% of visitors purchased |
| Drop-off / abandonment | Where people leave a flow | 40% abandon on the shipping step |
| Retention | % who come back after a period | Day-30 retention: 22% |
| Feature adoption | % of users who use a feature | 15% of users created a saved search |
Attitudinal metrics
Attitudinal metrics capture how people feel and what they think, through standardised questionnaires.
System Usability Scale (SUS)
Created by John Brooke in 1986, SUS is a 10-question survey with a 5-point agreement scale ("I thought the system was easy to use"). Answers are converted into a score from 0 to 100. A score around 68 is considered average; above about 80 is generally excellent.
SUS is quick, free and reliable even with small samples, which is why it's the most widely used usability questionnaire. Note that the score is not a percentage.
Single Ease Question (SEQ)
Asked right after a task: "Overall, how difficult or easy was this task?" on a 7-point scale. Simple and good for comparing tasks.
Customer Satisfaction (CSAT)
"How satisfied were you with…?" usually on a 1–5 scale, asked after a specific interaction like a support chat or a delivery.
Net Promoter Score (NPS)
"How likely are you to recommend us to a friend or colleague?" on a 0–10 scale. 9–10 are promoters, 7–8 passives, 0–6 detractors. NPS = % promoters − % detractors (range −100 to +100).
NPS measures loyalty and overall sentiment, not usability. It's affected by price, brand and support as much as by design.
Google's HEART framework
Google's research team created HEART to help product teams choose user-centred metrics at scale. It has five categories:
| Category | Measures | Example metric |
|---|---|---|
| Happiness | Attitudes and satisfaction | SUS, CSAT, app ratings |
| Engagement | Depth and frequency of use | Sessions per user per week |
| Adoption | New users of the product or a feature | % of new users who create a project in week 1 |
| Retention | Users who keep coming back | % active again after 30 days |
| Task success | Efficiency, effectiveness, errors | Completion rate, time on task |
Goals → Signals → Metrics
HEART is used with a simple process to make sure metrics connect to what you're actually trying to achieve:
- 1Goal — what does success look like for users? e.g. "New freelancers send their first invoice quickly and confidently."
- 2Signal — what behaviour or attitude would show we're succeeding? e.g. "They complete the first invoice without abandoning."
- 3Metric — how do we measure that signal? e.g. "% of new users who send an invoice within 24 hours of sign-up", and "median time to first invoice".
Common mistake: Measuring what's easy instead of what matters. Page views and time on site are easy to collect but often misleading — a user spending ten minutes on a page might be engaged or completely lost.
Funnels and analytics
A funnel shows how many people move through each step of a flow. It quickly shows where the biggest drop-offs are, so you know where to focus research.
Product page 10,000 ████████████████████ 100%
Add to cart 3,200 ██████▍ 32%
Checkout started 2,100 ████▏ 21%
Shipping details 1,260 ██▌ 12.6% ← biggest relative drop (−40%)
Payment 1,050 ██ 10.5%
Order complete 900 █▊ 9%Analytics tells you where and how many. Qualitative research tells you why. When you see a drop-off, follow it with session recordings, usability tests or a short on-page survey.
A/B testing
In an A/B test, users are randomly split between two versions — the current design (A) and a variation (B) — and you compare a metric such as conversion rate. Because the groups are random, a significant difference can be attributed to the design change.
- Test one main change at a time so you know what caused the difference.
- Define the primary metric and the sample size before starting.
- Run the test long enough to cover weekly patterns and reach statistical significance — don't stop as soon as B looks better.
- Watch guardrail metrics (e.g. refunds, support contacts) so a win on one metric isn't a loss elsewhere.
- A/B tests need traffic. Low-traffic products may learn more from qualitative testing.
Benchmarking and reporting
Measure before and after a change, and track key metrics over time. A simple UX scorecard for each core task keeps progress visible:
| Task | Success | Avg. time | SEQ (1–7) | Trend |
|---|---|---|---|---|
| Create first invoice | 91% | 1m 42s | 6.1 | ▲ improved |
| Add a client | 96% | 0m 38s | 6.5 | ■ stable |
| Export for taxes | 64% | 4m 05s | 4.2 | ▼ needs work |
Key takeaways
- Behavioural metrics show what people do; attitudinal metrics show how they feel — use both.
- SUS gives a 0–100 usability score where about 68 is average; NPS measures loyalty, not usability.
- HEART (Happiness, Engagement, Adoption, Retention, Task success) with Goals → Signals → Metrics keeps metrics meaningful.
- Funnels show where people drop off; research explains why.
- A/B tests compare versions with random groups — test one change and wait for significance.
Exercise
Define UX metrics for a product
- Choose a product (real or imagined) and one key user goal.
- Fill in Goals → Signals → Metrics for two HEART categories.
- Sketch the funnel for the main flow and guess where the biggest drop-off is.
- Write the research you'd do to find out why people drop off there.
- Describe one A/B test you'd run, with its primary and guardrail metrics.
Test yourself
Measuring UX quiz
5 questions. Score 70% or more to complete the lesson.
