TelcoApp

Mobile App · UX Case Study

TelcoApp

One intelligent control centre for every SIM a Pakistani user carries.

Scroll, it starts with a balance that vanished

The summary flips. Go on, tap it.

How it went
The problemPeople juggle 2 to 3 SIMs with no view of where their balance goes.
The researchInterviews and a survey; four user segments came out of it.
The insightUsers don't want data. They want decisions.
The betForecast when data runs out, then suggest the right plan.
The guardrailAutomation stays off until the user turns it on, and every switch can be undone.
The review4 heuristic findings, 3 fixed in the screens.
NextA moderated test with 5 multi-SIM users.
Chapter 1 · Context and stakes

Three SIMs, and no idea where the balance goes.

Almost every Pakistani mobile user runs two or three SIMs. Each operator has its own app, bundles change constantly, and none of them tells you when something is about to go wrong. Balance disappears and people simply recharge again.

Left alone, that means money lost to the wrong bundles, surprise depletion mid-day, and a lasting distrust of every telecom app on the phone.

“My data just disappears. I have no idea how. I just keep recharging.”

Interview participant
8user interviews (5 everyday, 3 power users)
126survey responses
5competitor apps analysed
4user segments found
Chapter 2 · Role and boundaries

Research first, then the screens.

I owned
  • Research plan, interviews and survey
  • Segmentation, personas, empathy and journey mapping
  • Competitive analysis and problem framing
  • Information architecture and the seven product modules
  • Interaction design, UI and the UI guide
Boundaries
  • Concept project: no operator data integrations or engineering build
  • AI behaviour is designed at the interface level; models were not trained

What the research showed

Four segments, one shared feeling. Across all of them the complaints were the same: people did not know where their balance went, could not compare packages, and were never warned before something bad happened.

User segments from the survey

Every 100 people who answered the survey, by segment. Hover a segment.

Tap a card to flip it ↻

The journey, and how it feels

Tap a point on the curve.

1. Awareness2. Search3. Decision4. Use5. Recharge6. Reflection

“Where did my balance go?”

Frustration
Pain point

No visibility into consumption

Opportunity

Clear visual usage tracking

Competitor analysis
AppStrengthWeaknessOpportunity
Jazz WorldPopular, stableConfusing bundles, poor AIAI-driven auto-plans
MyZongBetter UI than othersNo predictive toolsDeep AI usage forecasting
MyTelenorClean UINo multi-network supportUniversal SIM management
Ufone AppSimpleLacks marketplaceDevices and add-ons in one place
SimosaTrue cross-network supportNo AI, forecasting or optimisationIntelligence on top of cross-network
My Verizon / T-MobileStrong AI, dashboardsLocked to a single operatorFirst multi-operator AI app

“Users don't want data. They want decisions.”

Chapter 3 · Key decisions and trade-offs

Users don't want data. They want decisions.

Decision 1 of 5

Recommend decisions instead of reporting usage

Interviews: “I wish the app would just tell me the best package.”

Options on the table
Why this one

Every segment, including power users, asked for an answer, not more charts. A recommendation also serves Khalid, who will never read a graph.

What I gave up

Recommendations have to be trusted, so each one shows the reason and the saving. That costs screen space on every card.

Try it

Try it: when does your data run out?

AI Pulse moves the warning to when you can still act calmly. Set your own numbers and see the forecast and the suggestion it would make.

AI Pulse · Estimated

Your data runs out in 11 days

TodayRuns outRenews

A working sketch of the forecast idea using simple averages. The concept marks forecasts from balance messages as estimates.

What ships first

Seven modules came out of research. A first release needs three: SIM onboarding, the dashboard with AI Pulse, and the Plan Optimizer. They answer the two biggest complaints, "where did my balance go" and "which plan should I pick". The shop, add-ons marketplace, billing and concierge depend on operator partnerships and come later.

7 modules came out of research
01

AI onboarding

  • Auto-detects SIMs
  • Learns usage patterns
  • Sets personal preferences
Ships first
02

Adaptive dashboard

  • Real-time usage bars
  • Predicted bill and exhaustion
  • AI Pulse micro-recommendations
Ships first
03

AI Plan Optimizer

  • Analyses past 30–90 days
  • Simulates next month
  • Shows the saving, offers auto-optimise
Ships first
04

Device shop

  • Smart cards and EMI calculator
  • AR previews
  • AI device comparison
Later
05

Add-ons marketplace

  • Personalised add-ons
  • One-tap activation
  • Auto-add when needed
Later
06

Billing & payments

  • Predictive billing
  • EMI tracking
  • Auto-pay with rules
Later
07

AI concierge

  • Conversational assistant
  • Activates roaming, buys add-ons
  • Links into the right screen
Later
The core flow: a forecast on the home screen, a suggestion that shows its reasoning, automation that stays off until chosen, and a switch that can be undone from Home until it happens.
The states that make it hold up: an operator that shares no live data, a new SIM with too little history, no connection, and notice before any automatic switch.
UI guide: Inter, a blue primary and a mint secondary.
Chapter 4 · Outcome and evidence

A concept, checked hard.

TelcoApp is a concept, so there is no launch data. The evidence is the research: every screen traces back to an interview finding or a survey segment.

Heuristic review and walkthrough

I reviewed every screen against Nielsen's 10 usability heuristics, then walked through the four test tasks below step by step as a first-time user. Findings are rated by how likely they are to stop someone completing a task.

4 findings · 3 fixed in the screens · 1 open

HighOpt-in sheet · consistencyFixed

A toggle and a "Turn on" button sat together, so it wasn't clear which one actually turned automation on.

Removed the toggle. Two buttons: "Turn on auto-optimise" and "Just this switch".

HighWalkthrough · undo the switchFixed

Once you leave the confirmation screen, Home didn't show the scheduled switch, so the undo was hard to find again.

Home now shows a "Scheduled · 1 Nov" card with Undo until the switch happens (screen 5).

LowHome · affordanceFixed

"Recharge" looked like one of the SIM filter chips.

Styled as a button.

MediumTab bar · recognitionOpen

Only the active tab shows a label. Plans, Shop and Bills are icons alone, which new users have to guess.

Show short labels on every tab, or at least until the user has opened each one.

How I'll test it next

A moderated test with 5 multi-SIM users on a clickable prototype of the core flow.

  1. Task 1Find out when your data will run outPasses if: 4 of 5 give the day without help
  2. Task 2Decide whether to take the Zong suggestionPasses if: 4 of 5 can say why it was suggested
  3. Task 3Set up automatic switching with a Rs 1,500 limitPasses if: No one turns it on by accident
  4. Task 4Undo the switchPasses if: 5 of 5 within 30 seconds

Results and the changes they lead to will be added here.