Case study · Verizon Prepaid
How the Lifecycle Playbook took monthly churn from 7% to 3% on a ~$2B prepaid wireless business — improving 10–15% quarter over quarter, nine quarters straight, measured against holdout the whole way. Every number here is protected revenue, not projected.
Prepaid wireless is lifecycle marketing on hard mode. No contract holds anyone. Customers pay month to month, often in cash, and can walk to a competitor at any register in America. The business: about four million customers and roughly $2B in revenue at 60%+ margin — with monthly churn running at 7%, which in prepaid arithmetic means the average customer stays a little over a year. The mandate was one word: down. And behind it, the revolving door spinning in both directions — acquisition pushing customers in, churn pulling them out. And the starting point was thinner than any org chart admits: no plan, no playbook — one autopay campaign nobody could prove was working, and a three-week wait every time a question went to the central data team. The assignment wasn’t to run campaigns. It was to slow the door.
Stage 01 · Coverage
Before any model, any segment, any redesign: the ability to see. With answers three weeks away, we built measurement inside the team — Excel and a weekly readout at first, unglamorous on purpose, then near-realtime dashboards. Only then, funnel monitoring at every gate of the customer journey. The discipline is unglamorous — instrument the moments that matter and ask, at each one, what behavior proves this customer is receiving value. The monitoring earned its keep fast. Pricing launched a new $30 value card — five dollars cheaper than the plan it replaced, with the data allowance trimmed to make the math work. On a spreadsheet, a sharper price point. In a Walmart parking lot, it was a customer who had just paid for a smartphone plan that couldn’t open email, WhatsApp, or FaceTime — with no wifi in reach even to find out why. Churn on the cohort shot up. The call centers lit up. And the funnel monitoring put a number on what was walking out through the payment funnel: $3M a month.
The fix was fast once the leak was visible — affected customers got the $40 plan at the $30 price, and the funnel healed. The lesson outlived the fix: the leak wasn’t in any message. It was in a decision made three teams away, and only journey-level instrumentation could see it. Not a strategy problem. A silence problem. Someone just had to look.
Stage 02 · Relevance
Behavioral segmentation sorted four million customers onto the three paths that matter — progressing, stuck, or leaving — and every program targeted one path with one job. The clearest example: payment-failure churn wasn’t one problem. Some customers had left. Others were loyal people in temporary financial distress being treated like bad actors. So we built a save program for exactly that segment. The biggest hurdle wasn’t the data — it was the sentence “we tried this before and it failed.” They were right. They had tried. It had failed — because it had been run as a collections initiative, not a customer one. Done right, it took coordinated changes across care, sales, collections, risk, product, UX, and payments, all landing at nearly the same time. No single team could have shipped it; that is exactly the point of a lifecycle approach. It retained 2,500 customers in its first month and turned a collections moment into a loyalty moment.
Stage 03 · Intelligence
With coverage complete and segments proven, predictive analytics and AI-driven personalization earned their place: real-time decisioning choosing the next best action for each customer in real time, with rules constraining the models and humans arbitrating the conflicts. When the churn model, the cross-sell model, and the adoption model all wanted the same customer’s attention, an arbitration layer decided what deserved it — because you cannot shout four things at one person. That’s what “One Size Fits You” actually takes.
Stage 04 · Value
Every major journey ran with a control group. These are incremental results — what happened that would not have happened without the program.
Personalized journeys lifted customer lifetime value 33% — average lifetime stretched from 18 months to 24, because customers who get a relevant experience stay for it. App adoption rose 25% along the way — not because an app launch was promoted, but because app usage was one of the few behaviors that predicted survival. Promote the behaviors that create value. Not the feature that shipped Tuesday. Same acquisition spend — six more months of every customer.
Epilogue · What happened next
When Verizon acquired TracFone, the Prepaid playbook scaled to the combined portfolio — 16 brands, 22 million customers, a $10B business — with 300+ next-best-action models built from zero across lifecycle stage, channel, plan tier, and behavior. It survived a five-fold scale-up because it was never about the size of the base. It was about the order of operations: coverage, relevance, intelligence, value.
You don’t need four million customers for any of this to work — the playbook is about decisions, not headcount. Walk the journey. Find the silences. Sort customers onto the three paths. Add intelligence only where it changes a decision. Measure the only way a CFO accepts. The physics are identical at 20,000 customers; the tooling is just cheaper now.