Unifying 100M customers into one app they actually wanted to open

Verizon had two apps. One for Wireless. One for Fios. Both served the same household, often the same person, and neither knew the other existed. Only 27 percent of multi-service households understood they could manage Fios in the Wireless app. The app rating was 2.8 stars. Bill payment took 4.2 minutes, twice the industry standard. Thirty-four percent of negative reviews cited the same problem: promotional banners and upsells were burying the three tasks that accounted for 86 percent of sessions. Check bill. Make payment. View data. The app was built around what the business wanted to say. The job was to rebuild it around what the customer came to do.

I ran 32 user interviews and analyzed 2,400 app reviews before opening a design tool. The diagnosis came fast.

The app had an identity problem, not a usability problem. Two divisions had built two experiences with two navigation models, two billing flows, and two visual languages. A customer with both Wireless and Fios was managing two mental models for one account. The fragmentation was organizational, not technical. The app was a mirror of the org chart.

86 percent of sessions were three tasks. Check bill. Make payment. View data usage. Everything else, the promotions, the upsells, the feature marketing, was furniture blocking the door. The app was optimized for the 14 percent of use cases the business cared about, not the 86 percent the customer came for.

Promotions were not just noise, they were actively destructive. One research participant said it plainly: "I just need to pay my bill before it is late. Where do I do that?" That question, from someone who had already downloaded and opened the app, was the indictment. The app had succeeded at acquisition and failed at the one job that mattered after acquisition.

The app had problems everywhere: navigation, billing, account management, promotions, notifications, service visibility, performance. Five designers and a six-month window. The question was not what to fix. It was what to fix first, and what to leave for later.

I prioritized by working backward from the 2,400 app reviews. Negative reviews clustered around three themes in a clear hierarchy: billing confusion (41% of complaints), promotional clutter (34%), and service fragmentation (25%). The temptation was to start with service unification because it was the most architecturally ambitious and the most compelling internal narrative, "one app, one experience." But the data said billing was where the pain was sharpest. A customer who cannot pay their bill does not care whether their services are unified.

Ship order: billing first, then home screen, then unification. We rebuilt the billing experience in weeks 5–10, which gave us a working surface we could put in front of users and show measurable improvement on the highest-complaint category. That early win bought organizational credibility for the harder structural work. When I went to the Wireless and Fios leadership teams to argue for the unified account model, which required both divisions to give up control of their separate experiences, I had data showing the billing redesign had already moved the needle. The request was not "trust my instinct." The request was "look at what happened when we fixed billing, and let me apply the same approach to the rest."

The promotional layer was deliberately last. It was the most politically sensitive surface. Marketing had budget authority over the app's promotional real estate. I needed the billing and unification wins in hand before I could make the case that moving promotions would improve engagement, not reduce it. Sequencing the political risk after the credibility-building work was as much a part of the design strategy as the interface itself.

The 2-day design sprint produced four concept directions. We tested all four with 12 users before converging.

Direction 1: Service-switcher model. A single app with a toggle at the top, Wireless or Fios, that swapped the entire interface. Two apps wearing one coat. Users understood the toggle but resented it. One participant: "Why am I choosing? You know I have both." The switcher made the organizational seam visible to the customer. That was the opposite of the goal.

Direction 2: Dashboard-first with cards. A home screen of modular cards (bill summary, data usage, Fios status, promotions) that the customer could rearrange. Tested well with power users. Failed with the 86 percent who came for one task. The customization was a feature for people who enjoyed configuring apps. The target user wanted to pay their bill and leave.

Direction 3: Chat-first interface. A conversational UI where the customer typed or spoke what they needed. "Pay my bill." "Show my data." Ahead of its time in 2019. Speech recognition for account management was not reliable enough, and the fallback to traditional UI when the chat failed created a worse experience than starting with traditional UI.

Direction 4: Task-aware home with unified account. Surface the customer's most likely next action based on account state. One login, one billing view, one service dashboard. No switcher, no customization, no chat. Just the three tasks that mattered, presented in priority order, with everything else one tap away. This was the direction we shipped.

Directions 2 and 4 were the final two. The deciding factor was the 86 percent number. A dashboard-first model optimizes for browsing. A task-aware model optimizes for doing. The data said the customers were doers.

I set four design principles publicly, before the team drew anything, and held the work accountable to them throughout.

Tasks first, promotions second. The home screen would surface the customer's top tasks, not the business's top campaigns. Promotions would earn space through relevance, not through org-chart authority. This was the fight. Marketing owned significant real estate in the existing app. The argument I made was that the 2.8-star rating was the cost of that ownership.

One account, one experience. A customer with both Wireless and Fios would see one login, one home screen, one billing interface. The two-app mental model would collapse into a single personalized view. This required cross-divisional alignment that did not exist when the project started.

Progressive disclosure over feature density. Surface the 20 percent of functionality that serves 80 percent of needs. Everything else is one tap away, not zero taps away. The instinct in enterprise apps is to show everything. The data said showing everything was why bill payment took 4.2 minutes.

Clarity over cleverness. Plain language. Transparent billing. No dark patterns around plan changes or add-ons. The previous app had been optimized to create confusion that drove support calls. We would measure success by whether support calls went down.

The redesigned app shipped across iOS and Android over a six-month build with a phased rollout. The work was not a reskin. It was a structural rearchitecture of how the app understood the customer.

The home screen became task-aware. It surfaced the customer's most likely next action based on their account state: upcoming bill, recent data usage, pending order. The promotional layer moved from the home screen to a dedicated discovery surface where it could be relevant without blocking the primary workflow.

The billing experience was rebuilt from scratch. A single interface managed both Fios and Wireless billing, surfaced upcoming charges with plain-language breakdowns, and reduced the payment flow to the minimum viable steps. The previous flow had been designed to show the customer more information at every step. The new flow was designed to show them less, better.

The unified account view consolidated Mobile, 5G Home, and Fios into a single personalized dashboard. Plan changes, device management, and service updates all lived in one place regardless of which division owned the underlying product. The customer did not need to know which division they were interacting with. That was our problem, not theirs.

We ran a 2-day design sprint early in the process, tested 6 concept directions with 12 users, and went through 4 rounds of iteration before the phased rollout. During rollout, we A/B tested aggressively. One finding reshaped the final product: less intrusive promotions did not reduce conversion. They improved it by 23 percent. The data validated the design principle the business had been most skeptical of.

Four rounds of iteration between the design sprint and the phased rollout. Each round changed the product.

Round 1: Task hierarchy. The first prototype surfaced bill summary, data usage, and account status equally. User testing showed a clear priority order: bill first, always. Customers' eyes went to the dollar amount before anything else. We restructured the home screen to lead with billing state (upcoming amount, autopay status, days until due) and moved data usage and account management below. Time-to-bill-pay dropped from 3 taps to 1.

Round 2: Promotion placement. We tested three positions for promotional content: inline between task cards, bottom of home screen, and a separate discovery tab. Inline promotions performed worst on task completion time and had the highest frustration scores. The discovery tab performed best on both task completion and, unexpectedly, on promotion engagement. Users who chose to browse promotions spent 3x longer on promotional content than users who were forced to scroll past them. Less visibility, more engagement. This was the data point that won the internal argument with marketing.

Round 3: Unified billing display. The first version of the combined Wireless + Fios bill showed both charges in a single total. Customers with both services panicked because the number was higher than either bill alone, and the breakdown was not immediately visible. We added a service-level breakdown with expandable line items. Bill-related support inquiries in the test group dropped 26% from the previous round.

Round 4: Phased rollout tuning. During the 6-week rollout, the A/B data surfaced a device-specific pattern. Android users completed bill payment 15% faster than iOS users on the same flow. Investigation revealed the Android back button created a natural "undo" safety net that made users more confident in tapping forward. We added an explicit back/edit affordance to the iOS flow. The gap closed within two weeks.

The app went from 2.8 to 4.6 stars within 90 days of launch. Bill payment dropped from 4.2 minutes to 1.8 minutes, a 52 percent reduction. Billing-related support calls fell 31 percent. Monthly active users grew from 8.1 million to 10.4 million. Multi-service adoption, the metric that mattered most to the business, increased 47 percent. Net Promoter Score improved 18 points.

The number that mattered most to me was the promotions finding. The business had protected promotional real estate in the app for years on the assumption that visibility drove conversion. The data showed the opposite. When promotions moved out of the way and into a context where they were relevant, conversion went up. The instinct that marketing reach equals marketing effectiveness did not survive contact with what the customer actually wanted.

The thing I would do differently is fight harder for the unified experience earlier. The two-app problem was obvious from the first week of research, but I spent the first two months building consensus across divisions instead of shipping a prototype that demonstrated the unified model. In retrospect, the prototype would have done the consensus-building for me. The lesson: in large organizations, a working demo that shows the customer's experience is a more persuasive argument than a deck that describes it. I have started every cross-org alignment conversation with a prototype since.