Exploring facial recognition in doorbell cameras
How quick concept testing and 10 minutes of discovery increased stakeholder buy-in and clarified the feature’s value.
Employer
Vivint
Year
2025
Project background
The AI team was developing the facial recognition model for the doorbell camera. This was one of the company’s first customer-facing AI features.
I partnered with a designer to build a prototype that we could test with both Vivint customers and customers of competing products. The initial research focused on testing two core interactions within the concept. However, I convinced the PM to extend the interviews slightly so we could include a brief discovery section at the beginning.
Kickoff
I initially met with the PM and designer to better understand the project. They walked me through the work completed so far and the questions they hoped the research would answer.
I reviewed their questions and dug deeper into how they planned to use the research findings. They explained that the research would help build the case for the feature’s value with leadership. At that point, I suggested adding a brief discovery section to the beginning of each interview to better understand where customers might find value in the feature. The team agreed.
Concept work done by the designer and PM
Prototype screens showing well-known celebrities to make communication about the concept easier than using strangers.
Moderated Interviews
I conducted 12 moderated interviews, with the first portion of each interview focused on discovery. All participants owned doorbell cameras: half were Vivint customers, and half used competing products.
Key interview objectives:
Understand the scenarios in which people currently use their doorbell cameras.
Test the two core tasks: tagging a person and merging people.
Explore where users would expect to find the experience within the app.
Deliverables, Impact, and Insight
The designer attended most of the interviews, and we synthesized the findings together. I then created a research deck for the PM that addressed each key objective and highlighted participant quotes, pain points, and opportunities. The findings gave the team enough evidence to secure leadership support and move the feature forward.
Additionally, grounding the discovery questions in a tangible concept surfaced three broader themes. These themes continued to emerge as we conducted research on additional outdoor and doorbell camera features, ultimately giving the teams three north stars for desired UX outcomes.
