Customer analysis on Image Injection
BrowserStack · Live • Sole researcher (E2E) • Product revamp research
Context
BrowserStack Live lets QAs test web and mobile apps across real cloud devices. Image injection is the feature that simulates a device's camera, letting a tester inject an image (like a QR code) where the device would otherwise see the physical world. It's the test surface for any flow that depends on a camera: scanning, identity verification, document capture.
The product team suspected this feature was underutilised relative to its strategic value. I led the customer analysis to find out why.
What I owned
End-to-end customer analysis on a single feature, anchored on a customer journey mapping exercise. Five user interviews. I owned the work from screening to prioritisation.
- Screening. Drafted the screener, recruited 5 QA professionals across experience levels and team sizes.
- Script. Designed a task-based usability study built around a real-world scenario (testing a QR-code scan on a cloud device).
- Interviews. Ran all 5 sessions, observed task completion, probed for friction at every step.
- Journey mapping and synthesis. Built a customer journey map of the image injection flow, rated friction at each stage across all five participants, distilled the output into themes.
- Prioritisation. Walked PM through the findings and jointly prioritised what was worth fixing first.
I transitioned off the project before implementation. The recommendations are with the team.
Method: customer journey mapping
The core analytical move was a customer journey map. Interviews alone surface anecdotes; a journey map turns them into a structured artifact a PM can act on.
- Decomposed the flow into seven stages, from session setup all the way through task completion.
- Rated friction at each stage for each participant on a Low / Medium / High / Fail scale, using behavioural signals from the recording (time-to-progress, recovery attempts, verbal confusion).
- Aggregated into a heatmap so the high-friction stages were visible at a glance, with an overall journey rating per participant for cross-user comparison.
- Layered the themes on top. Each high-friction stage mapped to one or more of the four themes below, so the journey wasn't just a diagnosis, it was a treatment plan.
The heatmap is what shifted the conversation with PM. Anecdotes get debated; a coloured matrix of where every user struggled, stage by stage, doesn't.
| Journey stage | P1 | P2 | P3 | P4 | P5 |
|---|---|---|---|---|---|
| Stage 1 | Low | Low | Medium | Low | Medium |
| Stage 2 | Low | Low | Low | Medium | Medium |
| Stage 3 | High | High | High | High | High |
| Stage 4 | High | High | High | High | High |
| Stage 5 | Medium | n/a | High | High | High |
| Stage 6 | Medium | High | High | Fail | Fail |
| Stage 7 | Low | Low | Fail | Fail | Medium |
| Overall | Low | Low | Medium | Medium | Medium |
- Low friction
- Medium
- High
- Fail
Participant labels anonymised, stage names hidden, internal ratings scrubbed. The shape and rigour of the artifact are intact; the proprietary detail is not.
The objectives
The study was scoped around six questions:
- Discoverability. Can users find the feature?
- Concept understanding. Do users understand what image injection does, vs. file upload?
- Mental model. Where do users expect camera-related controls to live?
- Usability. How do users actually attempt a camera-based test, and where do they stall?
- Debugging and error handling. What happens when they pick the wrong tool, or it fails silently?
- Improvement signals. What would make this workflow more intuitive?
What I found
Four themes emerged, each tied to specific product behaviours. Together they accounted for almost every high-friction stage on the heatmap above. Available on request.
Action items
This section contains the prioritised list of recommendations that came out of the study. Available on request.
Strategic investment areas
Beyond the tactical UX fixes, the work identified feature-level capability investments needed to unlock enterprise adoption (parity, friction, scope). Available on request.
Outcome
- Made the strategic case. The study moved image injection from an underused surface to a recognised high-potential investment area. The journey heatmap and themes were the artifact that got the conversation there.
- Translated qualitative research into a prioritised, sequenced plan. I left PM with a ranked backlog (impact vs. effort), not a list of issues to debate. The plan moved into the upcoming roadmap cycle largely intact.
- Built an artifact the team could act on without me. The heatmap and themes are durable. After I transitioned off, the team had what they needed to execute against the plan rather than re-interpreting raw interview data.