Security AI
MobileGrowth Strategy
Enhancing Support to Reduce Returns and Churn
MobileDesign System
We had just shipped a shiny new app experience—and somewhere along the way, help got lost in the hallway. Returns crept toward ~30%, and reviews were loud: people felt abandoned the moment something broke.
This wasn't a “tweak the FAQ” moment. It was a trust-and-survival moment. The north star became simple to say and hard to do: turn support from a deal-breaker into a competitive advantage—so customers could exhale, get unstuck, and stay.
What follows is the messy middle: the audit, the arguments, the pixels, and the metrics that proved we weren't just rearranging buttons.
Launched
Where we landed: a Support Center that actually feels like a front door—search, plain-language tags, and routes in from the places people already panic (onboarding, Wi-Fi, offline devices).
Project overview
Fair warning: bad support UX is a team sport. I led design, but the wins below belong to PM, CX, iOS, Android, and web engineers who debated edge cases with me until the coffee went cold.
My Team
3 Product Designers, Product Manager, iOS Developer, Android Developer, Web Developer.
My Role
Lead Product Designer.
Responsibilities
Product Strategy, Stakeholder Presentations, Project Planning, UX Research.
Timeline
Nov – Feb 2025.
The Problem
Migration day is exciting in the roadmap deck. In the wild? Users suddenly couldn't answer a basic question: “Where do I even get help?” The old mental model didn't map to the new surfaces—and every dead-end felt like the product was shrugging at them.
The numbers told the same story in a sharper voice: roughly three in tenpurchases bounced back as returns. That's not a copy tweak; that's a signal that confidence broke before the hardware did.
So we zoomed out. The job wasn't to bolt on a chat widget—it was to rebuild continuity between “something went wrong” and “someone (or something) has my back.”
What I Accomplished
If you only read a résumé bullet, you miss the glue work. Here's what actually moved the project forward—usually with sticky notes, not applause.
- Held the redesign end to end with CX, PM, and engineering—so “support” wasn't a side quest, it was part of the same product story.
- Turned noise into narratives by stitching reviews, tickets, and return data into a storyline leadership could act on (not just frown at).
- Facilitated the messy alignment—workshops, trade-offs, and “what ships first” decisions that made the roadmap honest.
- Coached newer designers & writers on tone and patterns—because a design system only matters if people actually enjoy using it.
Understand
What are we hearing?
Step one was embarrassingly simple: listen without defending. After the migration, the app didn't just look different—it behaved different. Help paths people memorized simply… vanished.
The return curve was the rude awakening. When ~30% of buyerssend hardware back, you're not debating button color—you're debating whether the product feels safe to own.
Bring a Smile to Our Users
Cheesy name, serious intent. We wanted people to feel relief, not rerouted—the kind of calm you get when the product says, “I've got you,” instead of “good luck out there.” Fewer returns and less churn were the scoreboard; dignity in the moment of failure was the real design brief.
Aashish B. (1 star): “Support has been completely useless and technically clueless every single time. This has been one of the worst support experiences I've ever had.”
Kathleen R. (1 star): “Do not invest in Arlo security cameras. You will NEVER be able to contact a live person in customer service.”
Alex M. (1 star): “It's impossible to reach anyone from customer service. This is the worst company I have ever dealt with.”
Entry Point Audit
Here's a fun fact about enterprise-y apps: support isn't one door—it's thirty doors, half of them disguised as error states. We mapped 30+ entry points into the Support Center so we could stop arguing from memory and start arguing from facts.
Why bother? Because inconsistent entry points quietly train people to give up. If “help” only exists where one PM remembered to link it, you don't have a product—you have a scavenger hunt.
Blue vs. orange dots = different routing realities (MOT vs. RouteThis). Translation: the spreadsheet was our shared brain.
Three support pillars
Users didn't fail because they're lazy—they failed because we asked them to guess the right channel under stress. So we stopped pretending one catch-all screen could hug every intent. Instead, we framed help around three honest jobs: chat when you want speed, phone when you want a human, and KB when you want to self-serve without shame.
Live agent chat
For the “I can type while pacing” energy—async, traceable, and less awkward than crying on hold.
Phone support
For the messy stuff—returns, accounts, “this camera hates my router” moments where tone of voice matters.
KB articles
For the proud troubleshooters (and 3am fixers) who want answers tied to the exact error, not a generic pamphlet.
Final Outcomes
Designs & Prototypes
If you skip straight to pixels, you miss the plot. The UI below is the outcome—but the plot is: we earned the right to show those options at all.
Research & Findings
Early interviews had a recurring punchline: people weren't mad at bugs—they were mad at feeling marooned. The app had “support” in the org chart, but not in the journey.
So we pressure-tested what “good” could mean, and kept landing on the same trio: live agent chat, human phone support, and KB articles that don't read like a ransom note. Not because three is a magic number—because those three map to how real humans recover from panic.
Research Finding Part 1
The delightful surprise? When we surfaced the right options clearly, people actually used them—no bribery required. Clarity beat cleverness.
Before
Support Center
Support Requests
Front Door Ultra — Internet connection
Support for Your Products
- Front Door Ultra
- Home SmartHub
- Subscription
After
Support
We are here to help.
Support by Phone
Talk to an Arlo Expert. We'll give you a call to help sort everything out.
Call a Live Agent →
Support by Chat
Start a Chat →
Self-Help Made Simple
Visit the Support Article →
Support by Phone flow
Phone flows are where UX goes to prove it has manners. We weren't designing a form—we were designing a promise: “we heard you, we're coming, here's something useful while you wait.”
So yes: request → confirm → keep people oriented with relevant articles instead of leaving them staring at a spinner. Small copy choices; huge trust payoff.
1 · Support overview
Support
We are here to help.
Support by Phone
Talk to an Arlo Expert. We'll give you a call…
Call a Live Agent
Support by Chat
Start a Chat
Self-Help Made Simple
Visit the Support Article
2 · Request call
Support by Phone
Talk with our next available agent.
Enter your phone number, and our support agent will be with you shortly.
Country
🇺🇸 +1
Phone Number
612-555-1234
Estimated Wait Time: 2 Minutes
Request Call
3 · Confirmation
Support by Phone
Your phone call request has been received.
Our support agent will be with you shortly.
Estimated Wait Time: 2 Minutes
While You Wait
- The Camera failed to connect error appears… ↗
- How do I mount my Arlo Essential Outdoor Camera… ↗
Browse Support Articles
Continue While Waiting
Cancel Request
Results
We knocked it out of the park on all levels!
I'll be honest—I don't love chest-thumping. But I dolove when the metrics rhyme with the story we told in critique. Here's the short version: we made help easier to find, kinder to use, and less expensive for the business to clean up after.
Translation for busy readers: fewer tickets, more meaningful taps, and a system that won't fall apart the next time we ship fast.
- Successful redesign— We reshaped layout and flow so iOS and Android users weren't playing two different games of hide-and-seek with the same features.
- Improved usability— We carved friction out of the “I'm stuck” paths, and ticket volume responded: ~25% fewer customer support tickets. That's not just calmer users—it's calmer ops.
- Addressed user needs — We shipped what people had been asking for—loudly—and engagement followed: ~40% more interaction with key in-app features. Turns out when you listen, the product feels alive.
- Cross-platform consistency— One story, two platforms: validated UX so Android and iOS didn't drift into accidental cousins.
- Design system created— We left behind components and patterns so the next team doesn't have to reinvent the wheel—or argue about spacing at midnight.
