Virgin Media O2 rebuilds its data stack and speeds up the customer experience

When was the last time you upgraded your phone? Like many people now, you probably don’t rush to buy the latest model each year. That’s because today’s devices and network speeds are already fast enough to meet most everyday needs.

Virgin Media O2, a telecommunications company, noticed this shift in customer behavior as well. In order to grow, Virgin Media O2 realized it needed a new playbook: one that prioritizes a standout customer experience (and not just selling new phones).

To stay competitive, the telco giant needed to anticipate customer needs, personalize interactions, and deliver value quickly with every interaction. But to do that, Virgin Media O2 had to face another big problem: twenty years of technical debt.

Cutting through two decades of technical debt

Because of the Virgin Media and O2 merger, the company’s data lived in two on-prem systems. Both systems were siloed, entrenched in technical debt, and required inefficient workarounds to support daily operations.

Nowhere was this more painful than in the call center. Customer support agents had to switch between 8-10 screens across two different systems, just to answer a single customer question. For customers, this led to long wait times and, predictably, poor Net Promoter Scores (NPS).

To deliver a better customer experience, Virgin Media O2 had to address its technical debt and modernize. But rather than try to lift and shift these legacy systems to the cloud, its newly formed digital data team made a big bet: take the Greenfield Approach and start over from scratch.

It was a bold move, but it paid off. In just a year and with four engineers, the team implemented entirely new standards for data infrastructure inspired by the Toyota Production System:

• Stability first: products are designed for consistency and expected to continually improve.
• Standardization: every workflow, process, and data pipeline follows the same format.
• Stop at the point of failure: if something breaks early in the data pipeline, stop processing the data immediately to identify the root cause. Resolve the issue and then continue with clean data.
• Just-in-time data delivery: data is delivered with consistent latency, at the right moment for a specific purpose.
• Customer-centric innovation: team members are encouraged to innovate and build thoughtfully.

Powering continuous data delivery with dbt

Implementing these standards was made possible with dbt Labs. By making the dbt platform a critical part of its stack, Virgin Media O2 has transformed how its data is modeled, tested, orchestrated, and delivered:

• Modular, incremental data models: the team rebuilt their pipelines as modular models that update incrementally. This significantly reduced compute load and increased pipeline efficiency.
• Reduced waste with delta-tracking models: with dbt’s support for incremental logic, delta-tracking, and self-referencing models, the data team cut processing time from 47 minutes to less than 1 minute. It’s a dramatic improvement that aligns with call-center SLAs.

• Continual data flows: the team used dbt triggers to create a pull-based system. Once the initial step of a model succeeds, the next step begins. Now the team manages 500 runs per day, compared to a single batch job before.
• Automated quality checks: to ensure data integrity, the team uses dbt’s built-in testing framework to run over 600 automated data quality checks per day.
• GDPR compliance: data governance is embedded right into the architecture’s foundation. If the entire stack is wiped, the team can rebuild it in under eight hours with one click—honoring GDPR compliance requirements.

Today, the team can deploy to production within 24-48 hours. It’s the pace required to innovate and win customer love, and it’s a major competitive advantage.t

Happier customers, higher NPS, and cost savings

With a modern data stack in place, the data team swiftly built a unified call-center dashboard.

Now when a customer call comes in, support agents can pull the customer’s information in less than 60 seconds. They can verify the customer’s identity, review account details, and provide upsells based on ML-driven recommendations, without switching between systems.

Already, the dashboard has had a powerful impact. Customers are spending less time on the phone, and agents are helping more customers than before.

The results are evident in Virgin Media’s NPS, too, which improved by 26%. Overall customer satisfaction is up by 3% and call center efficiency increased by 3%. That may seem small, but it’s a significant change that has generated strong cost savings.

A customer-first transformation that scales

In the race to become customer-first, Virgin Media O2 is setting the pace. By integrating best practices from manufacturing and software engineering, Virgin Media O2 has redefined data standards for telco. With dbt, it’s ready to scale with AI and continuously raise the bar for the customer experience.

If you're a data leader at an enterprise and you’re thinking about how to modernize your data stack, we can help. Talk to our team to build your data strategy, or sign up to connect your warehouse and start building.