Joshua Watkins, enterprise network automation developer at Cox Communications, and the entire Cox Communications networking team recently faced a significant challenge: more effectively managing tens of thousands of networked devices across all its business units.

If the network automation team got things right, Cox Communications would be better equipped for digital transformation success in the years ahead. If they failed to meet the challenge, the telecommunications services provider would face increased risks of outages and business disruptions.

In recent years, Cox Communications has made significant investments in its digital transformation efforts, including $19 billion in network and digital product investments over the past decade. During that time, Cox saw its digital sales rise to 50% of total sales, up from 25%. The telecommunications provider has also increased its time-to-market for products and improved its customers’ online experiences. 

For Cox, a substantial part of its overall digital transformation strategy focused on improving its internal network automation. The telecommunications industry faces mounting pressure to reduce operational expenses while simultaneously improving service quality and accelerating deployment timelines. For Cox, network automation addresses these challenges by eliminating human error, reducing manual intervention, and enabling consistent policy enforcement across all business units.

For large enterprises, the cost of one hour of downtime ranges from $100,000 to more than $300,000. For some vertical markets, such as financial services, that hour of downtime could cost between $1 million and $5 million every hour. Comprehensive network automation not only mitigates such costly outages but also enables sustainable, rapid, and resilient application and service rollouts, allowing for more rapid entry into new business opportunities, as well as improved productivity and customer experiences. 

Those businesses that don’t automate their networks risk higher network costs and poor operational agility and responsiveness to market changes. 

That’s not news to Watkins. He and the Cox network automation team recently transformed the management of their network operations manual efforts to automated network management so that they could more effectively manage upwards of 80,000 network devices. “We defined a new operational baseline and decided that [automation] was how we were going to manage the network going forward,” said Watkins.

Cox’s Network Automation Transformation

Cox’s journey from manual network management to comprehensive network automation, or “network-as-code,” illustrates the mounting pressures faced by not only large-scale service providers but also how the strategic implementation of network automation, combined with an internal cultural transformation, can dramatically improve availability, agility, and operational efficiency.

Before embracing “network-as-code,” Cox Communications faced significant operational challenges that threatened the integrity of its networks and operational efficiency. For instance, manual configuration changes during late-night maintenance windows complicated operations and created unpredictable variations across Cox’s extensive network infrastructure.

Essentially, network-as-code is the application of software development principles—such as automation, version control, and continuous integration—to network management. This enables network administrators to define, deploy, and manage network configurations using machine-readable files rather than relying on manual processes.

Before these automation efforts, Cox’s network management relied heavily on individual engineers maintaining configuration settings, scattered spreadsheets, and inconsistent documentation rather than centralized systems. As Watkins explained, when network configuration information is stored in someone’s memory or on a napkin, dangerous single points of failure are created. And when those engineers change roles or employment, crucial systems information moves along with them.

Cox found itself managing tens of thousands of network devices and an increasingly complex network without a unified understanding of their current state or intended network configurations. This lack of visibility made it extremely difficult to assess compliance with security policies, respond effectively to emerging threats, or implement consistent changes across the network.

Watkins further explained that subtle, untracked changes accumulated over time and that this “configuration drift” increased the chances of future network outages and disruptions. “It’s just a future incident waiting to happen,” Watkins said. 

Network Automation is Critical for Digital Transformation

The Cox team recognized that successful network automation relied on high-quality data, and by gaining control over a wide range of network data, they could achieve comprehensive network control and subsequently automate network management processes more effectively. Cox centered its network data management efforts on Nautobot, a solution by Network to Code. Using Nautobot’s Single Source of Truth (SSoT) application to centralize device data, network topology, and connectivity and configuration data. This enabled the Cox network automation team to normalize data from aging devices that lack APIs, implement golden configurations for standardization, and force configuration compliance across their network. 

With DiffSync, a specialized library for bidirectional data synchronization, Cox’s network engineering team was enabled to integrate multiple internal systems, including Cisco Catalyst Center, Meraki, and SolarWinds, to gather all inventory and device information centrally into Nautobot. 

Today, while not all aspects of Cox’s network management are automated, the level of network automation they have reached to date improves network resilience. “There are a lot of nuances in the network that we still have to account for, and network automation is an ongoing journey of improvement and a focus on engineer-driven approaches,” Watkins said. 

Watkins advised teams embarking on enterprise network automation transformations to prioritize establishing comprehensive network inventories and baseline configuration standards before implementing automation tools. As Watkins added, attempting to automate poorly understood or inconsistently configured networks would cause additional complexity and trouble rather than the operational benefits sought.

The Future of Network Automation

By 2026, Gartner estimates that 30% of enterprises will automate more than half of their network activities, representing a dramatic increase from under 10% in 2023. This reflects both technological advancement and the recognition of automation’s business benefits. 

For instance, advanced AI-driven systems enable proactive network management through continuous analysis of network traffic behaviors, statistical heuristics, and machine learning models that can automatically detect and remediate misconfigurations and network errors, as well as flag unusual traffic patterns to prevent service degradation. These systems will also better forecast potential network failures and optimize power consumption by dynamically changing device settings. 

Where will network-as-code and network automation lead to in the years ahead? The machine learning algorithms that analyze historical data within the network, as well as code architectures, will better predict potential failures and automatically implement corrective actions before issues impact users and customers.

The global network automation market is projected to grow from $24 billion in 2024 to $104 billion by 2033, according to the market research firm IMARC Group. This growth reflects the recognition that manual network management won’t scale to meet modern operational demands driven by increasing network traffic, cloud infrastructure adoption, rapid deployment of connected devices, and the network demands from the growing adoption of AI and machine learning technologies. 

Those enterprises that embrace network-as-code and eventually self-healing networks will find that they hit scalability limitations, which restrict business growth. As their networks grow to accommodate new services, locations, or acquisitions, manual management approaches will require proportional increases in staffing. That constraint will force those organizations to either accept increased operational risk and business disruptions or significantly increase operational expenses to maintain service levels. 

Those that do move to network-as-code and network automation will deploy new applications and services much more rapidly than their lagging competitors with manual operations and will reallocate engineering talent from maintenance to innovation, thereby accelerating other digital transformation initiatives. That acceleration will create a competitive advantage through faster time-to-market for new offerings and more responsive adaptation to business demands.