ISZ GROUP
Results at a Glance
| Outcome | Impact |
|---|---|
| Business value generated over four years | €300M (incl. €100M in 2024) |
| Business days saved through process intelligence | 500 |
| Improvement in customer retention | 15% |
| Reduction in data errors via GenAI preprocessing | From 20% → under 4% |
A Fiber Network Is Only as Good as the Process Behind It
Deploying fiber broadband at national scale is an enormously complex operation. For a major European telecom provider — one of the continent’s largest, with revenues exceeding €17 billion — the challenge wasn’t the technology itself. The fiber was capable. The infrastructure teams were experienced. The market demand was clear.
The problem was process.
The provider operates through eleven regional divisions, each responsible for fiber deployment across its territory. Some regions managed rollouts internally; others relied on certified partners. Each had developed its own workflows, timelines, and quality standards over years of independent operation. With over one hundred discrete activities involved in a single business fiber deployment, these regional variations had compounded into a web of inconsistency that was silently draining efficiency and frustrating customers.
Deployments that should have taken weeks stretched into months. Customer installations that should have been completed in a single visit frequently required callbacks. And when leadership asked a simple question — “How long does it actually take to deploy fiber in Region 6 versus Region 3?” — nobody could answer with confidence, because the data was fragmented across systems that didn’t speak to each other.
The Limits of Intuition-Based Optimization
The provider had already invested significantly in operational improvement. Regional managers had their own ideas about what was causing delays, and many of those ideas were reasonable. But without empirical process data, optimization efforts were based on experience and intuition rather than evidence.
The Head of Automation captured the fundamental shift needed: process improvement must be grounded in facts. Not in what teams believe is efficient, but in what the data proves is 12%, 13%, or 14% more efficient — and then systematically building on those gains.
This realization led the provider to ISZ GROUP.
Deploying ISZ Prism: Starting Small, Proving Fast
The provider selected ISZ Prism — ISZ GROUP’s process intelligence platform — to map, analyze, and optimize its fiber deployment operations. The approach was deliberately incremental: start with a single region, prove the methodology, then scale.
Phase 1: Single-Region Discovery
A dedicated process analyst embedded within the provider’s fiber team used ISZ Prism to ingest operational data from one region’s deployment workflow. The platform automatically reconstructed the end-to-end process — from initial customer order through provisioning, scheduling, installation, and activation — revealing the actual sequence of activities, their duration, and their interdependencies.
What emerged was strikingly different from the assumed process. Activities that teams believed were sequential were actually running in parallel — but with no coordination, creating bottlenecks at handoff points. Certain steps that were supposed to take hours were consistently taking days, but only in specific scenarios that hadn’t been visible in aggregate reporting.
ISZ Prism’s variant analysis identified the specific process paths that correlated with on-time delivery and those that predicted delays — giving the team an evidence-based playbook for intervention.
Phase 2: National Expansion
With the single-region proof point established, the initiative expanded to all eleven regions. ISZ Prism was configured to accommodate each region’s operational nuances — integrating both legacy workflows and the newly optimized processes from Phase 1. The platform’s ability to normalize data across heterogeneous source systems was critical; without it, cross-regional comparison would have been impossible.
The rollout followed a structured cadence: bi-weekly workshops with regional process owners to review findings, address deviations, and share practices that were working. This bottom-up approach — empowering the people closest to the work to act on data-driven insights — was essential to driving adoption.
By the end of the national rollout, the process intelligence initiative had saved 500 business days across the organization. These weren’t theoretical savings — they were measured reductions in cycle time, elimination of redundant handoffs, and optimization of resource allocation across regions.
The lead process analyst summarized the approach: “We visualized the data, hosted workshops, and brought everyone to the table. Simply by identifying which activities should be managed internally versus externally made a transformative difference.”
From Process Visibility to Customer Experience
The operational gains were significant, but the most strategically important outcome was the impact on customer experience.
The “Once-and-Done” Target
One of the provider’s most ambitious goals was achieving a “once-and-done” installation model — completing fiber activation in a single customer interaction rather than requiring multiple visits. This had been a persistent pain point: customers who took time off work for an installation appointment, only to learn that a prerequisite step hadn’t been completed, were understandably frustrated.
ISZ Prism’s process intelligence revealed exactly which upstream activities most frequently caused installation failures, and at which points in the workflow intervention was most effective. By addressing these root causes systematically, the provider achieved single-visit completion on 60% of orders — with a clear trajectory toward the 90% target.
Data Accuracy as the Foundation
A critical enabler was the dramatic improvement in data accuracy. Before ISZ Prism, deployment tracking accuracy sat at just 13% — meaning that for 87% of deployments, the system’s record of what was happening didn’t match reality. Within weeks of implementation, accuracy climbed to 55% and continued improving as the process intelligence feedback loop matured.
This accuracy improvement wasn’t just an operational metric — it was the foundation for everything else. Accurate data enabled accurate SLAs, which enabled realistic customer commitments, which enabled trust.
Retention Impact
The cumulative effect of faster deployments, fewer failed installations, and more transparent communication drove a 15% improvement in customer retention — a metric that, for a provider of this scale, translates to hundreds of millions in lifetime customer value.
Scaling AI Across the Enterprise: €300M in Four Years
The fiber deployment project was transformative in its own right, but it was part of a broader automation and AI strategy that has generated over €300 million in business value over four years — including €100 million in 2024 alone.
Across the organization, ISZ GROUP’s platform now powers over 220 automations spanning network operations, customer service, billing, compliance, and field operations. The automation portfolio is growing at approximately 30% annually, with each new deployment building on the process intelligence foundation established by ISZ Prism.
GenAI-Powered Data Quality
One of the most impactful recent additions has been the application of generative AI to data preprocessing. Before automating a workflow, the provider now uses ISZ Nexus’s AI capabilities to clean, validate, and enrich the underlying data — reducing error rates from 20% to under 4% before automation even begins.
This sequence — intelligence first, automation second — has become a core principle of the provider’s transformation strategy. Automating a broken process at scale just produces broken outcomes faster. Fixing the process with data-driven intelligence, then automating it, produces compounding returns.
Document Intelligence for Billing and Contracts
The provider has also begun deploying ISZ Nexus’s document intelligence capabilities for automated bill and contract review. Using advanced OCR and contextual analysis, the system processes thousands of supplier invoices and customer contracts — identifying discrepancies, flagging compliance issues, and routing exceptions to the appropriate team — at a speed and accuracy that manual review could never match.
Principles for Telecom AI Transformation
This engagement reinforced several principles that apply broadly across the telecommunications industry:
-
Process intelligence must precede automation. Understanding how work actually flows — not how it’s supposed to flow — is the prerequisite for meaningful automation. Skipping this step is the most common reason telecom automation programs underdeliver.
-
Regional variation is a feature, not a bug. The goal isn’t to force identical processes across every region — it’s to understand which variations drive better outcomes and standardize around those. ISZ Prism’s variant analysis made this distinction possible.
-
Bottom-up adoption drives sustainability. The most sophisticated process intelligence platform is worthless if regional teams don’t trust and act on its insights. Bi-weekly workshops, shared ownership, and empowering local process owners were as important as the technology itself.
-
Customer experience is an operational outcome. Faster deployment, higher first-visit success rates, and accurate communication aren’t marketing initiatives — they’re operational capabilities that process intelligence directly enables.
-
Excellence in operations is excellence for customers. As the provider’s automation lead observed: “If you can deliver a service two or three days faster and make a customer happier, those are two or three days of additional revenue for the business.”
Case study published with client authorization. Identifying details have been modified for confidentiality. Contact ISZ GROUP for a detailed briefing on process intelligence and AI automation for telecommunications.
