Prove the value
Increase user awareness of app benefits and encourage proactive device cleaning.
Vodafone Secure Net: Making the Invisible Feel Powerful
Most security happens silently. This redesign made protection visible, actionable, and trustworthy. I led the design end-to-end — directing UX research, owning all decisions, and shipping across 4 markets.
Role
Lead Product Designer · sole design owner · directed dedicated UX researcher
Platform
Android
Timeline
4 months (research to rollout)
Reach
~100,000 users / 4 markets
Result
Drop-off 60% → 40% · NPS +1.5 · rolled out across UK, DE, RO, GR
The Problem No One Talked About
Security was working — but users couldn’t feel it
Most of what Secure Net does, it does silently. Threats are blocked in the background. Suspicious sites are filtered before users notice them.
That is great for security, but bad for perceived value. Users could not feel the product working, and when you cannot feel something, you stop trusting it.
The Cleaning Tool was the exception: scan device, find threats, remove them, see result. But almost nobody finished the flow.
Drop-off rate: 60%.
What We Found
Not confusion — unmet expectation
I ran user testing in the UK before changing the interface. Users started curious, then became uncertain and left.
The flow stacked orientation, virus-definition preparation, scan, and removal without telling people how long each beat would last. Timing swung with backend latency, device tier, and storage volume — so even reliable engineering felt arbitrary.
A six-step onboarding sequence still greeted people before the scan began — useful for education, expensive for momentum when someone only wanted to clean their phone.
UX Research
Understanding the drop-off, not just measuring it
I directed a dedicated UX researcher to run moderated sessions in the UK before any design work began. We observed 12 users attempting the Cleaning Tool flow on their own devices, tracking where hesitation turned into abandonment.
The sessions revealed that drop-off was not about confusion with the UI. Users understood what to do. They lost trust in when it would end. The flow gave no temporal anchors — no progress, no stage markers, no sense of how long each step would take.
We mapped the emotional arc: curiosity → uncertainty → impatience → abandonment. The inflection point was consistently around step 3 of 6, where onboarding content felt repetitive and the scan had not yet started.
Coming soon
Coming soon
The Design Decisions
Three coordinated decisions across UX, UI, and system behavior
Speed up the engine
Accelerate the scan process and surface relevant threat metadata earlier.
New design system
Adopt Vodafone Home Portfolio system for a more readable, accessible experience.
Design system upgrade became a UX redesign
The initial scope was a design system migration. During requirements review, I initiated research that reshaped the brief.
- Users wanted control over digital security and access to threat detail
- Making scans “educational and entertaining” was not validated by research
- Effort redirected toward clarity, speed, and trust
Three separate briefs. One design opportunity I shaped from research.
1. Collapse the flow
I removed the database update from the user path and moved it to a silent background pre-step. The experience became a clean two-step flow: Scan → Remove.
2. Set expectations visually
I redesigned screens to communicate progress clearly: active processing, threat detection, and clean-device confirmation became visually distinct states, not just text changes.
3. Improve backend communication
In parallel, I worked with engineering to optimize backend interaction, reducing real wait times and making the tool feel genuinely more responsive.
Execution screens split “work in flight”, “what we found”, and “what happened next” into distinct visual chapters instead of recycled layouts.
6 → 2
User-facing steps
−67%
Interaction steps
60 → 40%
Drop-off rate
Onboarding intro
Welcome screen explaining what Cleaning Tool does.
Education carousel
Six slides about threats, security tips, definitions.
DB update wait
Visible virus definition download blocking the flow.
Scan
Full-device scan with minimal progress feedback.
Threat list
Flat list of issues, no clear action hierarchy.
Completion
Generic success screen, no summary of what happened.
Scan → Remove
Two clear phases. DB update runs silently in background. Onboarding moved to an optional info touchpoint.
- Direct entry into scan with real-time progress
- Actionable threat ledger with severity and selection
- Earned closure: summary of cleaned items and time
- Education available on demand, never blocking
See how six disconnected steps became a focused two-phase experience.
Before and After
Old flow vs. redesigned flow
Drag the slider to compare the original and redesigned experience.
Design Exploration
Searching for the right visual language
Before converging on the final direction, I explored multiple approaches: onboarding tone, scan UI density, progress visualization, and completion states. The process included paper sketches, low-fidelity wireframes, and visual research across security and health-monitoring apps.
Research artifacts that informed the design direction -- scroll horizontally to explore.
Testing and Iteration
Tested, narrowed, iterated, and revalidated
I tested three design variations with users. One clearly outperformed the others, but moderated sessions exposed friction points that metrics alone would not reveal.
Mid-scan states were especially vocal: percentage motion helped, yet headline hierarchy, reassurance copy, and cancellation paths decided whether someone trusted the wait.
I iterated and retested until those points were resolved. Four months after the first research session in the UK, the redesign was live.
The Complete Flow
Every screen, every state, every edge case
The final design covers the full journey: dashboard entry, permissions, scanning with real-time progress, results with actionable threat management, cleaning with cancel/uninstall paths, and celebration closure. Scroll horizontally to explore.
Rollout and Results
Consistent performance across markets
The feature launched first in the UK (with baseline data), then in DE, RO, and GR where Cleaning Tool had not been available before.
UK (Q1 → Q3 2025): Drop-off 60% → 40%, Android app NPS 5.0 → 6.5.
DE, RO, GR: new market baseline with post-launch drop-off around 40%, consistent with UK’s improved performance.
Total reach: approximately 100,000 users across 4 markets.
Drop-off rate
60% → 40%
Android app NPS
5.0 → 6.5
What I Learned
Users judge reliability through communication
Security products often prioritize reliability over communication, but users experience both at the same time.
A flow can be technically reliable and still be abandoned if it feels broken. The biggest win here was not only simplifying steps — it was designing what users needed to feel, not just what they needed to do.