Network automation challenges are dampening success rates

Network automation technology promises operational efficiencies and reduced security risks for enterprise businesses, which makes it seem like an obvious investment area for IT leaders. Yet aligning data sources, adopting the right tools, and achieving success with network automation are daunting tasks that require IT teams to share data and work across domains to see results.

Research firm Enterprise Management Associates surveyed 354 IT professionals in January 2024 about their organization’s approaches to network automation. The results show that just 18% of surveyed IT professionals rate their network automation strategies as a complete success, while 54% noted achieving partial success, and 38% said they were uncertain of the level of success achieved or admitted failure with their network automation projects.

“Network automation continues to be a big problem and a big issue. Why does network automation continue to be so hard? Things continue to be stuck, and the problem isn’t going away,” said Shamus McGillicuddy, research director for the network management practice at EMA, who shared data from the firm’s recent report, Enterprise Network Automation: Emerging from the Dark Ages and Reaching Toward NetDevOps. “There may be no one path to network automation success, but there are a few things an organization can do to ensure success.”

The survey data reveals that most IT organizations are using a mix of do-it-yourself, homegrown tools, open-source software, and commercial automation technologies. According to EMAs’ recent webinar, network automation continues to challenge IT leaders who struggle to find the right skills on staff, to create a single source of truth about their environments, and to drive collaboration across IT domains.

Benefits of network automation

For those organizations seeing some success with network automation, the benefits are clear and more compelling with today’s complex multicloud environments. Two benefits topped the list of positive results of network automation investments: 33.9% of respondents pointed to operational efficiency (skilled personnel are more productive), and 32.8% of those surveyed cited reduced security risk. “In other words, network automation boosts network team productivity and eliminates errors that might lead to security events,” the EMA report says.

Other noted benefits of network automation initiatives include:

• Improved capacity management: 22.9%
• Network/application resiliency: 22.6%
• Accelerated incident response (e.g., mean time to insight, mean time to repair): 21.5%
• Increased/improved collaboration: 21.2%
• Agility (responsiveness to change and business needs): 18.1%
• Reduced regulatory compliance risk: 17.8%
• Empowering lower-skilled personnel to do more: 16.7%
• Capital expense avoidance (e.g., extending life of hardware, maximizing use of installed equipment): 16.7%
• Reduced configuration drift/improved design compliance: 16.7%
• Accelerated time to market for new applications/services: 16.7%

“At the scale we do things with thousands of locations, even doing a little thing like running a simple change or command on a network device is just a huge job,” said a network tools engineer at a Fortune 500 retailer in the EMA report. “Automation just makes life so much easier for people in terms of compliance, gathering config files from devices, running reports, and deployment.”

The EMA data also debunks the idea that network automation will in some way replace IT professionals. According to interviews EMA conducted with 10 anonymous IT leaders, the worry over automation is mitigated by the benefits automation brings.

“Automation takes human error out of the network. We’ve decreased outages by a significant amount,” said a network automation engineer at a large university in the EMA report. “When people hit a button to automate something, we have scripts running so many checks before a change is done. Also, we’ve empowered technicians who lack the skillsets to make changes and write configs. Field teams can be made up of people with lower skillsets, who can go out and hit a button to troubleshoot things.”

Despite the benefits reported by 18% of respondents, many business and technical challenges remain. According to the EMA report, IT leadership issues (buy-in, direction, commitment) were cited by more than 31% of IT leaders. More than one-fourth (26.8%) of the respondents pointed to staffing issues such as skills gaps and staff churn as a business challenge.

“The most challenging thing for me is the lack of network engineers who can contribute to automation,” said a network engineer at a midmarket business services company in the EMA report. “The community is small, and it’s hard to find people who can help you solve a problem.”

Another 25% pointed to budget issues as a pain point, and nearly one-fourth cited conflicts/collaboration issues between groups (e.g., NetOps, DevOps, SecOps). Security policy constraints were also reported by nearly 25% of respondents. Other challenges include difficulty with prioritizing specific use cases for automation (18.6%), unclear business value for investment (16.9%), stakeholder buy-in, engineers resist using automation (16.4%), and customer support limitations (16.4%).

Network automation technical challenges

Aside from the business hurdles in the way of a successful automation strategy, there are several technical challenges that must also be addressed. Difficulty integrating automation with other systems is cited as the biggest challenge, with more than 25% of respondents pointing to integration as a technical challenge. For nearly one-fourth of respondents (24.9%), network complexity/lack of standards were cited as one of the more challenging technical issues associated with network automation.

Other technical issues cited by EMA’s survey respondents were:

• Legacy infrastructure issues—lack of APIs, inconsistent features: 24.3%
• Tool complexity—automation is difficult to use: 23.7%
• Data issues (quality, conflicts, access)/fractured source of truth: 22.3%
• Tool stability—automation is buggy, unreliable: 21.5%’
• Inconsistent support across vendors, platforms, clouds etc.: 21.5%
• Legacy automation/technical debt—translation of old solution to new strategy: 21.2%
• Scalability issues: 20.1%
• Lack of certain features/functionalities: 17.8%

“Standardization is the biggest obstacle,” said a network tools engineer at a Fortune 500 retailer in the report. “When the network is not standardized and the data is not standardized and you don’t have a standard way of generating inventories and a source of truth, it’s a big problem. You can’t automate at scale because you’re forced to automate one device at a time without standardization.”