The Internet will tell you that automation is just minutes away from upending work in offices and industries, but if you spend enough time talking to tech executives and stakeholders, you will learn that it is not the silver bullet that execs hoped it would be, at least not at the moment.
Despite inflation and rising labor costs generating significant interest in process automation, and CIOs stretching their budgets to accommodate the costs, automation adoption rate (AAR) is dwindling at enterprises across America.
The reason is one – the trouble of building automation is too much to even consider for some, and others to maintain. A large percentage of automation projects get abandoned or shut down midway owing of the amount of work needed to see the real dividends.
The Barriers to Expanding Automation
Automation is great for getting small, repeatable jobs done. For example, it is a trusted tool for configuring periodic notifications, enabling email responses or processing employee payroll. Reliance on these systems is steadily growing across workplaces.
But it has been observed that automation initiatives like these are inclined to get stalled very quickly. Research shows that as enterprises graduate to take on more complex works, there are significant tailwinds to deploying the technology.
Complex jobs require clever and custom automation. That kind of automation only gets built once in a while when the need arises.
“It’s like spring cleaning in your house…You’re only motivated to do it once a year,” commented Tom Hollingsworth during the Delegate Roundtable at Networking Field Day Exclusive with Nokia in September that explores the reasons for stagnation of automation projects.
Another major stumbling block keeping enterprises from deploying automation is the network design. The data center fabric is simply not automatable by design.
“There’s a long laundry list of why network automation is not moving along faster than we think,” said Scott Robohn, networking and automation veteran. “We don’t really think about automation from a network design perspective. We don’t build it to be automatable.”
Scott also said that the lack of holistic understanding of how different workflows and processes chain together and how automation tools can affect change in them poses a big barrier to entry.
This is a growing reality in enterprises fueled by skill shortage in emerging technologies. “Most of the companies do not have anyone who knows automation on the staff,” pointed out Rita Younger, industry expert. “A big fear is if they hire someone trained up on automation who can realize the gains, how long it is going take to level up the next person.”
“We’re busy trying to keep the lights on, and keep the data center from going down,” she said.
The Faultline Separating Investments and Payoffs
Investments and returns are often misaligned in automation projects. Teams wind up overspending time and money building technologies without getting adequate value out of it. This difference has been a consistent friction point hindering automation projects in organizations.
“If people view automation primarily as a way to reduce effort and cut costs, they’re going to find that their cost go up in the short term because they have to do all the things they’re currently doing,” agrees VP of data center unit at Nokia, Michael Bushong.
Bushong believes that the true yardstick of measurement of results should instead be time. “When we equate automation to effort, we get one outcome, and when we equate it to time, we get a whole different outcome. People are oriented around effort for the most part, and I think they should be oriented around time.”
Measurable results come from a combination of investment, effort and time. Any new automation project, while likely to float up expenses initially, is bound to hit the TCO targets with sustained effort across time.
Unfortunately, the goals that professionals are left to pursue and realize at the end of these projects are set by C-suite executives steering the ship. Due to the nature of their work, they do not mind themselves with things happening in the lower tiers unless it impacts the earnings and opportunities. As a result, poor communication and fitful implementation are quite common with these projects. And when targets are not met, teams and departments are downsized as a way to recoup the losses.
A Hopeful Sign
For automation to be implemented at scale to make a real difference, companies must, first and foremost, build faith in the technology, said the panel. Across the globe, 30% to 50% RPA projects fail every year. This causes a lot of anxiety around relying on automation tools, causing practitioners to discard technologies when they break the first time.
Like great automation can future-proof operations, bad, underdeveloped ones can severely wreck the system and sink investments. The secret to building decent automation is hiring new skills, but also resetting expectations with clear, realistic goals in mind and approaching with an understanding that results will take time.
Bushong predicts that a workforce turnover will be the accelerant that kicks off adoption.
“Over the next 10 years, when network engineers age out and retire, the bumper crop of folks that follow are unlikely to come out of university with networking degrees and four and five letter certifications. If companies require crazy certification and rote memorization of vendor-specific syntax to be effective in an environment, they’re not going to be able to hire. The looming workforce transition will force people to make decisions if only because they have to hire.”
Unable to find new workers to do the job the old way, these companies will turn to automation for certain tasks. Meanwhile, the work to encourage and support talent upskilling must continue.
Parting Thoughts
Progress is incremental. Statistics show that in enterprises, some business units are implementing automation with more success, while others are yet to catch up.
Ron Westfall, research expert and analyst at The Futurum Group, has a positive outlook. “I think progress is being made. We’re getting more data, more evidence for that. For example, through network automation, you can decrease operational overhead by up to 40%. That is an encouraging stat.”
How an enterprise works toward exploiting the promise determines the outcome. “I’m not sure how many companies have deliberate plans, but it could be that the enlightened will have a plan, and for the others, it’ll happen and they may scramble to figure it out, but either way, you’ll see things will start to move. Individuals can decide whether that’s a painful move or a graceful leap,” said Bushong.
Change is afoot, and in organizations, the automation creep is getting real. Whether or not it kills jobs in scores, it will surely enforce real progress.
“We’re at the threshold of having breakthroughs that are going to be ecosystem-wide in terms of automation being implemented in a more widespread basis within data center environments,” Westfall concluded.
Catch the Delegate Roundtable conversation and other interesting presentations on network automation from Networking Field Day Exclusive with Nokia on the Tech Field Day website. Also check out Ron Westfall’s coverage from the event on The Futurum Group’s website.
