More than a quarter of US households now own at least one Internet-of-Things (IoT) device, and analysts expect the average home to host 10–15 connected gadgets by 2027. Yet these conveniences come with an awkward trade-off: microphones, cameras and sensors are always listening, always learning. 

Owners may grudgingly accept that bargain, but what about the friends, cleaners or Airbnb guests who never consented in the first place?

Prior work shows that “bystanders”—people who do not own a device but appear in its data—seldom realise smart appliances are collecting information about them (Apthorpe et al., 2019). Even when they do notice, the power dynamics of visiting someone’s home make it difficult to speak up (Yao et al., 2019). My co-authors and I wanted to know whether simple notifications—flashing lights, phone alerts, talking speakers—could close that awareness gap without ruining the social vibe of a dinner party.

What We Wanted to Find Out

Grounded in the privacy-notice design space, we asked three research questions:

• Which factors shape users’ and bystanders’ preferences for privacy notifications in smart homes?

• Which delivery channels—visual, auditory, mobile—do each group favour?

• How do those preferences shift across everyday situations such as working from home or hosting friends?

These questions recognise that “the smart home” is rarely a single-user environment; any solution must juggle multiple stakeholders with uneven control over devices (Thakkar et al., 2022) 

How We Studied 259 People Without Stepping Into Their Houses

We designed an online survey containing two mirrored versions: one for 136 smart-home owners (“users”) and another for 123 non-owners (“bystanders”). Recruitment ran on Prolific and included only adults based in the United States.

Participants first answered baseline questions about devices they own or have encountered, plus general privacy concerns. Then they evaluated four fictional—but technically feasible—notification mechanisms:

• Data Dashboard – a wall-mounted tablet showing detailed logs of every device on the home network.

• Mobility App – a smartphone app that scans the local Wi-Fi and pushes alerts when sensitive data leave the house.

• Ambient Light – a smart bulb whose colour and brightness signal the volume and security of outbound traffic.

• Privacy Speaker – a dedicated smart speaker that verbalises warnings such as “Your TV is streaming unencrypted data.”

• For each mechanism, respondents rated effectiveness, ease of use and comfort, then explained their reasoning in open text.

Finally, we presented four plausible scenarios—unlocking a smart door with biometrics, hosting a movie night, working from home and tracking health data—and asked which mechanism(s) they would want in that context. The mix of Likert-scale items and qualitative comments allowed us to quantify broad trends and still capture the nuanced social logic behind them.

What We Learned

1. One Size Does Not Fit All:  

Across the entire sample the Mobility App came out on top: 95 % of owners and 93 % of bystanders thought it would be effective, mainly because it delivers rich information privately to a device they already carry. Yet enthusiasm cooled in social settings; nobody wants to pull out their phone during a party just to decode flashing network packets.

2. Social Norms Trump Technical Detail:

Bystanders worried about looking rude if they touched someone else’s Data Dashboard—even though it held the most complete data logs. One respondent wrote that poking around a friend’s dashboard “would probably turn into a social faux-pas.” Owners, on the other hand, loved the transparency but disliked having to install another gadget on the wall. This tension illustrates how privacy mechanisms must satisfy etiquette as well as utility.

3. Unobtrusive Cues Are Great…Until They’re Not:

The colour-changing Ambient Light was praised for blending into the background, especially in work-from-home scenes where constant pop-ups would derail concentration. Yet both groups flagged two pitfalls: the meaning of red versus green is not obvious to guests, and subtle cues demand that people remain in the same room to notice them. Some bystanders even said a sudden red glow would be “kind of scary”.

4. Voice Alerts Divide Opinion:

Roughly four in five participants found the Privacy Speaker informative, but only half felt comfortable using it. Owners feared the device would blurt out sensitive details; bystanders worried the announcement would vanish before they could process it. Several said audio warnings would simply raise anxiety without offering any control .

5. Context Matters More Than Modality

When we mapped choices against scenarios, patterns emerged:

• Mobility App remained the most popular in every scenario, reinforcing its versatility.

• Data Dashboard ranked second overall but plummeted during a “visiting friend” scenario, suggesting that guests favour non-invasive methods.

• Ambient Light and Privacy Speaker gained traction in “work from home” where silent, lightweight cues minimise distraction.

Four Takeaways for Designers and Policymakers

Drawing on thematic analysis, the study proposes four design dimensions for future notification systems:

1. Easy Access: Information should be available from a personal device or a neutral interface so bystanders are not forced to handle the owner’s equipment.

2. Unobtrusive Modality: Visual or haptic cues that fade into the environment help maintain social flow, though designers must ensure clarity.

3. Home-Level View: People prefer dashboards that summarise all devices rather than piecemeal, device-specific alerts.

4. Actionable Controls: Awareness without agency can heighten anxiety; notifications should link to settings that let individuals restrict or delay data flows. 

What This Means—and Where We Go Next

Smart homes will never be zero-surveillance zones, but transparent, context-sensitive notifications can at least level the playing field between owners and accidental participants. Our findings suggest that hybrid solutions (e.g., a phone app that also triggers a gentle ambient cue) could balance richness, discretion and inclusivity.

Limitations remain: the study relied on hypothetical mechanisms and self-reported intentions rather than longitudinal behaviour. Future work should deploy working prototypes into real households and measure outcomes such as changed device settings or altered guest behaviour.

Conclusion

Privacy in connected homes is no longer just a technical problem; it is a social choreography involving multiple actors, devices and expectations. By surveying both owners and bystanders, our study reveals that the preferred choreography depends on who you are and what you are doing. A catch-all notification system does not exist—and that is precisely the point. Designers must account for context, etiquette and control if smart-home privacy is to feel like good manners rather than a high-tech faux pas.

Reference

[1] Apthorpe, N., Huang, D. Y., Reisman, D., Narayanan, A. & Feamster, N. (2019) ‘Keeping the smart home private with smart(er) IoT traffic shaping’, Proceedings on Privacy Enhancing Technologies, 2019(3), pp. 128-148.

[2] Thakkar, P., He, S., Xu, S., Huang, D. Y. & Yao, Y. (2022) ‘“It would probably turn into a social faux-pas”: Users’ and bystanders’ preferences of privacy awareness mechanisms in smart homes’, Proceedings of the CHI Conference on Human Factors in Computing Systems, New Orleans, USA, 13 pp.

[3] Yao, Y., Basdeo, J. R., Kaushik, S. & Wang, Y. (2019) ‘Defending my castle: A co-design study of privacy mechanisms for smart homes’, Proceedings of the CHI Conference on Human Factors in Computing Systems, Glasgow, UK.