Verizon turns every interaction into a personal experience

The Business Issue

Verizon did not start from scratch. It already ran one of the most sophisticated decisioning implementations in the world, delivering billions of decisions monthly across every touchpoint where a relevant decision could be made from website, app, call centers, retail stores, and outbound communications. Their platform was generating real business value and was deeply embedded in the wider Verizon ecosystem, so any change to it would widely affect ways of working, culture, and people.

But then the market shifted. In the first months of 2023, large language models made it possible to generate usable content at mass scale as well as the actions themselves, such as pick the best offer as the page opens, run it through approvals, generate the fulfillment codes, and show it as the page loads. Verizon realized that a few thousand offer and treatment combinations could become ten thousand to a million. No operating model absorbs that growth by expanding its teams tenfold.

The same shift shrank Verizon's competitive moat, since capabilities that had taken five or ten years to build could now be built in a couple of years. Incremental change was not enough. And rather than tune a platform that was still working, Verizon chose to leapfrog several generations at once and widen the moat far enough so it can’t be narrowed down by competitors.

The Solution

Most companies personalize by segmentation, in cohorts, eligibility groups, and audiences to target with certain offers. But Verizon set out to remove the segment altogether. Segment of Me is their journey to individual (like a fingerprint) hyper-personalization in real time. What a customer sees can change each day and each time they interact with Verizon, shaped by what they need in that moment rather than the group they were sorted into.

Pega had capabilities covering everything Verizon needed. The real decision point was how to manage cost at Verizon's scale, because deep learning, text generation, and image generation carry significant expense. The team anticipated the escalating AI cost curve well before it became an industry-wide conversation. Since Verizon already had an established AI platform that they had used for years, they combined the output from heavy AI work like deep inferencing, content, and image generation with Pega's adaptive models;[MC5.1] enabling real-time decisioning at scale. The result is an architecture that balances performance and cost to make hyper-personalization viable at Verizon's volume.

Yet technology alone does not drive success. At an organization the size of Verizon, spanning many groups with their own territories and priorities, organizational alignment proved the hardest part of the journey. They tackled it by bringing every stakeholder together from the beginning, spending two days working through what each group needed and what they wanted from the outcome, then aligning on a shared North Star. Regular cadences, readouts, and transparent communication followed, with obstacles and roadblocks named openly rather than avoided.

Rollout was equally deliberate. Rather than prove the platform on a quiet channel, Verizon started with its most “prominent, valuable placement,” that customers interacted with regularly to secure value early and build organizational momentum. The team rebuilt the entire proposition catalog and all back-end data connections, executed new implementation, and migrated traffic one percent at a time – with extensive measurement and testing at each step. This gave them a model and cadence that is replicable.

They then had to address Verizon's oldest implementation, which was also its most business-critical, created by people who have already retired. But rather than stop, Verizon kept one team on it while a second team operated in parallel – bringing new web, app, and contact center placements live, rotating people between the two, sustaining both momentum and morale.

The Results

Verizon completed the functional migration of all placements to its new hyper-personalization platform and celebrated the project as done at 26 months. The definition of “done” was determined up front and in writing, not simply as “install a new platform” but as business outcomes that ensure every placement, capability, and piece of functionality are live on the new platform. Moreover, several of the benchmarks Verizon set for system performance and overall business value are over-performing their goals, and the more advanced features are early in adoption.

The most durable outcome may be the team itself, battle-hardened by 26 months of it. Both Nanci Muriello and Travis McLaren received this year's Industry Visionary Award for Communications for the work. What comes next is already underway, with Segment of Me extending into agentic AI across Verizon's voice bot, chat bot, and digital channels. Because the hyper-personalized platform is already there, agentic will improve the conversational experience from the outset.

Want to dive deeper into Verizon's journey? Watch their PegaWorld session here.