Service Design: What It Is and How to Apply It
Service design optimizes the entire customer experience, not just the interface. Discover tools, principles, and how to apply it in digital environments.

UX research — or user experience research — is the discipline that studies real users to understand their needs, behaviors, motivations, and friction points when interacting with a digital product. It is not intuition. It is not what the CEO thinks the user wants. It is evidence.
The difference between a product that converts and one that doesn't usually comes down to the quality of prior research. According to Forrester, every dollar invested in UX research returns between 2 and 100 dollars, depending on the project phase. Detecting a problem during research costs 10x less than fixing it after launch.
UX research is not a luxury reserved for big companies. It is the most direct shortcut to stop guessing and start building on evidence. If you have ever launched a feature nobody used, you already know the cost of not researching.
UX research splits into two main families, and you need both:
| Aspect | Qualitative | Quantitative |
|---|---|---|
| Question it answers | Why? How? | How much? How many? |
| Typical sample | 5–15 users | 100+ users |
| Format | Interviews, usability tests, diary studies | Surveys, analytics, A/B tests |
| Output | Insights, patterns, verbatims | Metrics, percentages, significance |
| When to use | Exploration, discovery, qualitative validation | Measurement, prioritization, statistical validation |
The most common mistake is relying only on quantitative data (analytics tells you what happens, but not why) or only on qualitative data (5 interviews are not statistically representative). The combination is what drives solid decisions.
Semi-structured conversations of 30–60 minutes with real users. The most powerful method for uncovering motivations, fears, and decision processes that never show up in any dashboard.
When to use them: at the start of a project, when you need to understand user context, or when quantitative data reveals something you cannot explain.
Observing real users as they attempt to complete specific tasks on your product. You can run them moderated (you guide the user) or unmoderated (the user navigates alone, recorded on video).
When to use them: before launching a redesign, to validate prototypes, or when a page has a high drop-off rate and you don't know why. To complement with quantitative data, check out how to use heatmaps with Hotjar.
Users group and label content according to their mental model. It reveals how they expect to find information and is fundamental for designing navigation and information architecture.
Short surveys (3–5 questions) that appear at the relevant moment: after a purchase, when abandoning the cart, after using a feature. They capture quantifiable feedback in real context.
Users try to find items within a navigation structure without any visual interface. It validates whether the information architecture works before anything is designed.
Usability experts evaluate the product against established principles (Nielsen's heuristics, for example). It doesn't replace user research, but it is fast and catches obvious problems.
Users document their experience over days or weeks. This captures real usage in a natural context, including moments you cannot observe in a lab test.
Review of session recordings, heatmaps, funnels, and analytics data to identify behavioral patterns at scale. It bridges qualitative and quantitative research. Tools like Contentsquare are the benchmark here.
| Tool | Primary use | Price |
|---|---|---|
| Maze | Unmoderated tests, card sorting, tree testing | From 0 USD/month |
| UserTesting | Moderated and unmoderated tests, user panel | From 15,000 USD/year |
| Hotjar / Microsoft Clarity | Heatmaps, recordings, surveys | From 0 USD/month |
| Optimal Workshop | Card sorting, tree testing, first click | From 107 USD/month |
| Dovetail | Research repository, qualitative analysis | From 0 USD/month |
| Lookback | Remote moderated interviews | From 99 USD/month |
| Typeform / Google Forms | Quantitative surveys | From 0 USD/month |
You don't need all of them. With Maze (unmoderated tests), Hotjar (behavior), and a repository like Dovetail (analysis and storage), you can cover 80% of a product team's needs.
Exploratory interviews + competitive analysis + review of existing data. Goal: define real user problems, not internal assumptions.
Validation surveys + card sorting + tree testing. Goal: prioritize what to solve first and how to structure the solution.
Usability tests with prototypes + heuristic analysis. Goal: validate that the solution works before writing a single line of code.
A/B testing + contextual surveys + behavioral analysis. Goal: measure real impact and detect unforeseen issues. For more on this phase, read our guide on A/B testing.
Diary studies + follow-up interviews + retention metric analysis. Goal: understand ongoing usage and find improvement opportunities.
| Research investment | Savings / benefit | Ratio |
|---|---|---|
| 1 round of usability tests (5 users) | Detect 85% of usability problems | 10:1 |
| Continuous research program (quarterly) | 30–50% reduction in failed features | 20:1 |
| Pre-redesign research | Avoid a full redesign later | 50:1+ |
Jakob Nielsen demonstrated that with just 5 users you can detect 85% of usability problems. The cost of those 5 tests is a fraction of the cost of launching a feature nobody uses.
UX research is the foundation of any serious conversion optimization program. Without understanding the user, you are optimizing blind.
If you want to integrate user research into your conversion strategy, at Boost we combine research with experimentation. Get a free site diagnosis with Scan&Boost.
Adrià Vidal — Boost · Conversion Rate Optimization
Service design optimizes the entire customer experience, not just the interface. Discover tools, principles, and how to apply it in digital environments.