Telecoms operators have borrowed agile playbooks from software development for years, adopting sprint cycles and celebrating rapid deployments. But the industry faces a fundamental tension, as telcos must maintain 99.999% uptime and cover 95% or more of the population – expectations few other industries must fulfil. When a sprint delivers features at the cost of network stability, agility becomes operational debt arriving sooner. A faster release that triggers service degradation or increases incidents is not agility; it is failure.
The problem runs deeper than methodology. Networks are not software products. They demand different success metrics. Those operators that track progress using story points and burndown charts overlook what really counts – the effect on customers and the resilience of the infrastructure.
Operators must recognise that the next step in transforming telecoms must ideally focus on agile outcomes, defined by the ability to change networks, services and operating processes quickly while protecting critical components such as uptime, customer experience and infrastructure stability.
From sprint velocity to service outcomes
Traditional agile metrics answer an internal question: how quickly did the team deliver? Network operations need to answer a different question: what changed for the customer and the network?
Gartner’s research on CSP autonomous operations argues that communications service providers are pursuing autonomous operations to improve agility, efficiency, quality and customer experience. Gartner’s 2025 research also identifies network automation as a way to improve infrastructure agility and efficiency.
Here, the implication is important. A network team should not celebrate a shorter sprint if the resulting change increases incident rates or reduces availability.
Outcome-based iteration should instead connect engineering velocity to measures such as service availability and uptime, mean time to detect and restore service, customer-impacting incidents, change failure and rollback rates and service activation time. It also covers network capacity utilisation, automation coverage and successful autonomous actions, customer experience and complaint rates and revenue or service adoption affected by network changes. This does not mean abandoning software metrics. It means putting them in context.
Resilience emerges as a constraint on speed
Telecom networks are becoming more distributed, software-defined and complex. According to Analysys Mason, 71% of CSPs studied predicted they would be running cloud-native network functions by late 2023, up from 26% that were running such functions in December 2022. Disaggregation of network functions and microservices was also discussed.
Research also highlighted operational complexity created by disaggregated network functions and microservices. That complexity makes rapid change more valuable, but also more dangerous. A network change can propagate across multiple domains before engineers understand its consequences.
Gartner’s research on network digital twins makes a similar point, identifying that network teams struggle to test each software release and understand the systemwide effects of configuration changes. Digital twins can help automate testing of configurations, policies and equipment before changes reach production.
The lesson is straightforward: agility without assurance is simply faster risk propagation.
Measuring the value of each iteration
Agility based on outcomes demands that team members determine what a successful iteration is. For instance, an automation sprint for a network is not successful merely because engineers implemented a new process. The issues that matter are whether it reduces human intervention, enhances service quality, speeds up provisioning, or prevents customer impact.
In a network automation research report, Analysys Mason provided a useful illustration. Its analysis of optical network automation found that near-zero-touch provisioning reduced operational costs for provisioning by up to 81%, while operators reported increasing network capacity by 50% without additional engineering staff.
Those are outcome measures. They connect automation activity with operational capacity and business value.
Customer impact must enter the engineering loop
The strongest shift is to bring customer experience into network engineering decisions.
Forrester observed that telecom operators should connect customer experience data to financial and operational outcomes, including measures such as uptime, latency, incident rates, customer satisfaction and complaints.
That approach changes the conversation between product, engineering and operations teams. Instead of asking whether a feature shipped on schedule, teams can ask whether it improved the customer journey without creating unacceptable operational risk.
Forrester’s recent commentary on network reliability following the Verizon outage supports the argument. It noted that network automation, observability, AIOps, fast root cause analysis and overall optimisation play an essential role in ensuring network reliability.
Operating model needs to change
McKinsey’s analysis of agile telcos similarly emphasises organising around value rather than traditional organisational structures.
For network organisations, that means bringing engineering, operations, service assurance, customer experience and business teams around shared outcomes.
A practical model could include:
- Iteration: What network or service capability changed?
- Velocity: How quickly was it delivered?
- Stability: Did reliability improve, remain stable or deteriorate?
- Customer impact: Did customers experience better service?
- Recovery: If something failed, how quickly could the network detect and recover?
- Business value: Did the change improve revenue, capacity, cost or service adoption?
Agentic AI inflection point
Outcome-based agility gains urgency with autonomous systems. Gartner expects that by 2030, 50% of all organisations will deploy agentic NetOps with very little human intervention, compared with almost 0% in 2025. Such systems should be trained based on outcomes rather than delivery. An autonomous network agent optimising for sprint velocity would be catastrophic.
Agentic AI workflows will prevent a major outage autonomously, with tech leaders facing increasing pressure to improve resilience and uptime amid a growing and complex tech estate, according to Forrester’s 2026 predictions for technology infrastructure and operations.
Beyond software metrics
While the telecom industry does not need less agility, it requires a more mature definition of it.
IDC’s research on telecoms transformation notes that AI is moving into practical network operations, including predictive maintenance, automated network management, incident management and field support. IDC also points to the growing role of automation and orchestration as networks become more cloud-native.
Deloitte similarly describes AI-enabled network operations as a path from reactive to proactive maintenance, with the potential to reduce downtime and service degradation while making network operations more resilient.
The future of telecoms agility will therefore not be measured by how many tickets a team closes or how many story points it completes. It will be measured by how safely and consistently the network can evolve. Speed is an input, resilience is the outcome and for telecoms operators, only the combination creates meaningful agility.
Industry momentum towards outcome alignment
Operators are moving fast. Telecoms AI/GenAI investments are expected to soar from US$17.3 billion in 2024 to US$65 billion by 2029, according to IDC’s presentation on AI-powered networks in 2025. However, the investment is worthless without outcome-based operating models to guide it.
Forward-looking telcos embracing fibre rollouts, automation and cloud-ready architecture that is both modular and API driven can reduce IT costs by as much as 30% and boost average revenue per user (ARPU) by 10-15%, according to McKinsey’s research on critical bets for telco value creation. These gains only materialise when speed is tied to stability and measured against customer outcomes.
Speed without resilience is disruption. Agility without outcome alignment is chaos. Telecoms operators must move away from software metrics, adopt outcome-oriented velocity metrics and align agile approaches to network availability, service delivery and customer impact. Certainly, the future of telecoms operations lies with operators who will measure themselves based on customer experience, not merely sprints run.
Anna Ribeiro