Organizations stand to gain a lot by leveraging automation, especially the type that relies on self-learning and self-healing capabilities to further reduce human involvement. In fact, we’ve seen many companies decrease their organizations’ mundane IT activities by up to 70 percent after deploying effective and capable automation solutions. And yet, IT management is still largely managed by human employees. This begs the question: why hasn’t automation taken hold of this industry? The answer, it seems, is a combination of labor arbitrage, scripted automation and large vendor priorities.
While automation can effectively replace the many engineers required to manage an IT department, a number of organizations have historically depended more upon labor arbitrage because of its initial cost savings allure. The budget originally spent for local employees was diverted to outsourced employees, who could perform the same tasks at a fraction of the cost, effectively allowing enterprises to achieve the same outcomes for much lower prices.
However, labor arbitrage does not change how organizations operate or manage their IT environments—it just changes labor costs. Therefore, organizations with poor IT management systems have felt no incentive to update their systems or implement advanced automation. From their perspective, IT was handled. And the labor arbitrage companies providing these personnel had no incentive to drive automation either, as such a move would reduce their value and perhaps even eliminate jobs. Thus, labor arbitrage delayed organizations having to deal with their over-complicated, process-intensive IT environments.
With a rising cost of living in India, China and the Philippines, the cost efficiencies of labor arbitrage are narrowing for the most part; outsourcing providers are charging higher rates for their IT management services solely dependent upon human capital utilization. Labor automation, not task automation, is now challenging this approach.
The answer for many organizations unable to outsource has been to rely on scripted automation to manage their IT environments, following an “if A, then B” formula for automating tasks. For organizations that execute the same tasks on the same infrastructure time and again—for instance, rebooting the same servers at 6:00 a.m. every Tuesday—scripted automation is quite effective. But, for organizations with IT environments that are constantly changing, or those that are adopting more and more advanced technologies such as cloud computing, the benefits of scripted automation are much slimmer. For one, this type of automation requires significant upfront coding to script all of the actions that must occur in what order, and to program other settings or contingencies. This may take an engineer hours, or even days, and thus doesn’t save much time—especially if new scripts have to be developed frequently.
As a result of frustrating and inefficient scripted automation tools, many organizations adopted the mindset that automation is a lost cause, returning instead to expensive labor arbitrage or simply managing IT in-house with homemade, makeshift management tools. Automation solutions have actually evolved dramatically, and some now make use of autonomic qualities, meaning they can create automation solutions rapidly and accurately. After such a solution is uploaded into an organization’s IT system, it can track engineers’ reactions to various scenarios and mirror those in the future, duplicating the efforts of that engineer. For example, if there is performance degradation in one area of the network, the system will flag it to an engineer and “watch” the steps he goes through to test, diagnose and address this issue.
For similar problems in the future, the engineer won’t need to get involved and will simply be sent a notification that the system executed the fix and performance levels are now back to normal and strong. That means a more helpful solution, rather than a frustrating one.
The Other Side of the Equation
So why haven’t the world’s biggest IT service management providers brought to market such effective autonomic solutions? The companies that might fill this role of developing and marketing a better automation solution, in my opinion, have been conflicted. It’s a bit like a printer company that makes all its money on ink—why would they support a longer-lasting ink cartridge, or one that is more efficient? It may drive sales in the short term, but in the long term, it will result in reduced revenue. Similarly, many of the big automation players have partnered with labor arbitrage providers or built their own large offshore labor entities, and now push business back and forth between them as organizations change strategies. If they were to bring to market an automation solution that could remove the need for professional services or outsourcing, they would likely cut their own profits in the long run. This just wouldn’t be smart business.
Automation hasn’t gained momentum because most of the market participants have some sort of restriction in their business plans that limit their push for automation products. It feels a bit, to me, like the 2006 documentary, “Who Killed the Electric Car?”, which examined why electric cars weren’t introduced to the consumer market, despite being a better option over gasoline-powered vehicles. Luckily, that story has a happy ending, with many car manufacturers now prioritizing their electric models. Will the automation market end the same way, with efficient and effective solutions in the toolbox of every IT-intensive organization? I am hopeful that the answer is yes.