Connected devices live in real homes, so your research has to follow them there. The strongest iot user research methods are in-home contextual inquiry and multi-week diary studies, backed by setup and voice testing.
This guide is part of our wider look at iot user experience design, and it focuses on how to gather evidence.
| Item | Value |
|---|---|
| US households with smart home devices (2025) | 69.1 million |
| Owners who use only basic features | 40-50% |
| Connected devices abandoned within the first year | Roughly 15% |
| Qualitative sample | 8-12 households |
| Minimum diary study length | 14 days |
| Time per diary entry | Under 2 minutes |
These figures come from CleverX’s IoT research guide and its smart home diary guide. They can change, so check the sources before you rely on them.
Why connected devices need their own research approach
A connected thermostat is a dial, an app, a voice command and an automation rule at once, so testing a single screen misses most of the experience. Smart home adoption reached 69.1 million US households as of 2025, according to Statista figures cited in CleverX’s IoT research guide. Yet 40-50% of owners use only basic features, and roughly 15% of connected devices are abandoned within the first year, per Parks Associates data quoted there.
That gap between what a product can do and what people do with it is the real research target. The same guide explains what sets this category apart from ordinary product research:
- Many touchpoints: device, app, voice and physical controls, often inside one task.
- Many users: household members with different roles and different comfort with technology.
- Ambient interaction: passive automations and background monitoring that traditional usability testing never sees.
- Physical environment: WiFi coverage, device placement, room acoustics and home layout all shape the experience.
Start with in-home contextual inquiry
If you can run only one method, choose in-home contextual inquiry. CleverX calls it the highest-value IoT method because a lab cannot reproduce the physical layout, WiFi conditions, household dynamics and daily routines behind real usage.
During a visit, watch more than you ask. Note where devices sit, which rooms lose signal, who reaches for the app and who simply flips a switch.
When a visit is truly impossible, the same guide says you can simulate key elements in a lab: a multi-device ecosystem on a real WiFi network, competing devices and hubs, and realistic room configurations. Treat that as a fallback for narrow questions, not a replacement for the home.
Talk to everyone who lives there. The person who sets up the smart home is rarely the only person who uses it, as CleverX notes. Its smart home guide reports a common pattern: the person who set up the ecosystem loves it, and everyone else tolerates it.
Run diary studies long enough to see real habits
Diary studies capture what people actually do over time, rather than what they can do in one session. CleverX recommends diaries of 1-2 weeks minimum for tracking daily interaction patterns, automation usage and feature discovery.
The smart home diary guide sets the floor at 14 days, because it takes 5-7 days for novelty to fade and real usage patterns to emerge. It also says daily entries should take under 2 minutes, with secondary household members prompted 2-3 times per week.

A simple plan keeps the study manageable:
- Recruit households rather than individuals, and include at least one member besides the person who bought the devices.
- Write prompts that the least tech-savvy member can answer without help.
- Ask about devices used, anything that failed, and any automation created, changed or deleted.
- Plot entries per person over time and compare the setup person with everyone else.
Read the results in phases. In the diary guide’s framing, days 1-3 capture novelty behavior, so the interesting signal comes after that, when usage settles or stops.
Those moments of settling and dropping off are exactly what belongs in an iot customer journey map, where they become visible stages rather than anecdotes.
Sensors can strengthen the method too. A paper in the Proceedings of the Design Society, published in May 2024, proposes a diary method where IoT sensors trigger self-reports. Its authors report more diverse content and a better overview of users’ lives.
Test setup, voice and privacy without leading people
Setup is where many relationships with a device end, so give it dedicated sessions. CleverX recommends usability testing for onboarding, setup flows and app-device pairing. Our guide to iot device onboarding ux goes deeper on first-use flows.
Voice features need their own test, because speech varies by accent, dialect and background noise. Run voice sessions with a diverse group, and recreate a noisy kitchen or a television left on rather than a silent room.
Privacy is the trickiest topic. The guide advises against asking directly whether people are concerned, because participants give socially desirable answers. Instead, observe behavior, such as whether they cover smart camera lenses, and include every household member, since concerns often come from the person who did not choose to install the device.
How many households do you need?
Count households, not individuals. CleverX suggests 8-12 households for qualitative methods such as contextual inquiry and diary studies. Because each household may have 2-4 users, that yields 16-40 individual perspectives.
For surveys and A/B tests on companion apps, standard sample sizes apply, with 30+ for statistical significance. Voice testing needs 10-15 participants to capture the range of accents, dialects and speech patterns.
Frequently asked questions
Can you run IoT research remotely?
Show answer
Partly. According to CleverX, lab or remote sessions work for app navigation, settings configuration and data visualization. Anything that depends on the home itself, such as setup, automation creation or multi-room control, needs a visit.
Why not just interview the person who bought the devices?
Show answer
Because that one voice hides the rest of the household. The smart home guide says research that talks to one person in the household misses the point, since everyone else lives with the same product very differently.
What happens if diary prompts are too heavy?
Show answer
People stop answering. The diary guide warns that participation collapses within the first week when entries take longer than a couple of minutes. For members who are not comfortable with apps, it suggests simple paper journals.
How do you check what people really do versus what they say?
Show answer
Pair self-reports with device data. CleverX lists sensor log analysis as the method for comparing actual device usage with self-reported usage.
