Category: Systems-Automation

Working for many years in the commercial aviation industry, I witnessed profound changes in automation at almost every level. Ticketing counters gave way to self-service check-in kiosks. Passport control is dominated by document scanning, rather than human observation. Pilots manually control aircraft for only a few minutes at the beginning and end of a flight.

What will be the impact of the accelerated and the accelerating pace of automation across all sectors of the economy? Should we expect vast improvements in productivity and freedom from boring work, or should we fear threats to jobs, disruptions to organizations, and a strain on the social fabric?

In its November 2015 issue, McKinsey Quarterly presented a preliminary report on an ongoing study of automation in the workplace and reached two key conclusions:

• Very few occupations will be automated in their entirety in the short- and medium-term.
• However, certain activities are more likely to be automated, requiring entire business processes to be transformed and the jobs performed by people to be redefined – much like the bank teller’s role was redefined with the advent of ATMs 30 years ago.

The Magnitude and Impact of Automation

The study analyzed work activities for which current technologies could be deployed, and concluded that if automation were deployed in place of labor, its value in wages could amount for $2 trillion. Although we often think of automation as affecting primarily low-skill, low-wage roles, the analysis noted that even some of the highest-paid occupations (such as financial managers, physicians, and senior executives) perform a significant amount of activity that can be automated. The organizational and leadership implications of this conclusion suggest that organizations planning to take advantage of automation must implement significant process changes and even reengineer themselves – from the most senior positions to positons far down on the organizational chart.

Benefits of automation include labor savings, increased output, higher quality, improved reliability, and performing some tasks at speeds and at quality levels humans cannot. However, automation costs between three to 10 times what current manual processes do. The magnitude of the benefits and the costs required to attain the benefits of automation suggest that the ability to staff, manage, and lead increasingly automated organizations will become an important competitive differentiator.

This isn’t all-good or all-bad. Automation is our friend – think of E-ZPass at crowded highway tool booths. Automation is our bane – think of being chased through the maze of automated phone-answering systems, only to be looped back to the main menu.

According to the study,

• Approximately 45 percent of work activity could be automated to some extent, using current technology.
• Additionally, 13 percent of work activities could be automated, if we include emerging technologies that reliably understand natural language.
• Fewer than 5 percent of occupations can be entirely automated through current technology.
• However, approximately 60 percent of occupations could have 30 percent, or more, of their activities automated.

This means that automation is so far-reaching that it simply cannot be ignored. On another level, the rapid introduction of technologies based on artificial intelligence are challenging assumptions about what can really be automated. It is no longer the case that only routine, codifiable activities can be automated, and that activities requiring tacit knowledge or experience are difficult to translate into task specifications and so are immune from automation.

But beware. Automation is frequently oversold regarding achievable value and ease of implementation. It is almost always harder and takes more time than automation peddlers let on, and some \ expected benefits may be sacrificed in the process.  There is also the potential for “blow back,” where automation actually makes things worse. Sometimes, staff even passively sabotages automation efforts to avoid the need to adapt.

Bottom Line

Company leadership must monitor the speed and direction of automation and determine when and how much to invest in automation. Making these determinations will require leaders who understand the economics of automation and the trade-offs between augmenting versus replacing some activities with intelligent machines. Growing in a competitive environment frequently requires leveraging automation to innovate, and fully understanding the implications for human-skills development. In our technological world, choices about automation can both sharpen and cause the loss of a competitive edge.

I favor companies’ analyzing whether, and how, automation can create a competitive advantage. I also see this analysis as vital to a company’s overall strategy. In developing strategies and plans for clients I typically ask clients to think about “looking around the corner” to see what is coming in the future – and automation is one of the first places I probe.

Technology enables the automation of activities or services that accomplish specific functions or improve workflow. Business processes exist for many operational aspects of company activities, and leaders planning for their companies’ growth frequently ask me if they should consider automating some of those activities. My answer invariably is that it depends.

 

Generally, “whether to automate” needs to be broadened from a stand-alone question dealing with a single process to one considering it in the context of an organization’s overarching business strategy. Analysis must first determine the scope of the automation question.

 

The absolute first thing to avoid doing, though, is having a technical team determine if automation is the correct approach. Technical teams are great at answering the “how” question – how to technically do something – but not at analyzing if it should be done. While such teams are great at understanding how systems can solve problems, translating a technical solution to a business solution is typically not their strong suit. I’d advise bringing technical people into the conversation later, so that they can be told the business reasons for considering automation and the goals being considered – and then let them solve a precisely defined challenge.

 

Generally, companies struggle with the question of whether to start with a simple process or a larger process. The larger process could be more critical to develop first, which would offer the greatest value from automation in efficiency and savings. See the table below for a comparison between issues related to automating a simple process versus a larger, more complex process.

 

Comparison of Automation Efforts: Simple Versus Larger Process
Issue	Simple Process (Low-hanging fruit)	Larger Process  (More critical, more potential gain)
Success Speed	Faster – fewer steps	Slower – more steps, more integration
Progress	Quick win, but noticeable by fewer individuals	Slower win but more noticeable day-to-day because it touches many individuals  and has substantial, palpable impact
Chaos Control	Fewer people involved, easier to pay attention as issues emerge, then address them	Process automation time not properly estimated will generate doubts if it exceeds initial timelines and unexpected issues arise. Given the scope of the automation project, it might require extensive retraining and resistance by staff.
Process Importance	May not be considered critical in terms of cost savings, efficiency or improvements in customer relations	Larger, more important processes are generally deemed to be critical and likely to be given resources to improve upon them.
Return on Investment (ROI)	Lower	Generally, a higher ROI given the span of the process being automated

 

Here are some questions that I tend to ask that could help determine which processes offer the most value from automation:

 

1. Are there “paper heavy” processes? Is there a paper form (or e-mail or documents on a server) that gets routed to different participants as part of the process? Do your process workers waste time looking for forms or documents needed to complete a specific step?

 

2. Does the process require manual duplication of data, where something needs to be keyed in from one place to another?

 

3. Do processes “hang” because an individual was not alerted to proceed with the next step? Do other “routine,” time-consuming tasks halt the process in its tracks if the relevant employee is absent, overloaded with work or forgets?

 

4. Are there “repeatable labor” processes, in which individuals do almost the same thing in almost the same way in almost all cases?

 

If you answered “yes” to any of these questions, you are looking at a process with automation potential. If you’re still not  sure whether a process is a good fit for automation, or if you want more time to figure out the impact of such a “yes,” then it’s time to speak to the people who perform the process every day.

 

Understand though, that these process owners (people who do it now) might be comfortable with the status-quo; in fact, they might find that what they do today is ideal. It could be that many of them have already created workarounds to avoid potential process problems. Either way, these details need to be gathered to create an idealized automation scenario that reduces error rates and significantly decreases process cycle times.

 

To this point, you might have identified the benefit of automation, but what about other factors? Here are some that may lead you to not automate processes:

 

• Cost of Automation – What is the realistic ROI for automation? Include in this calculation the actual cost of automation, like software and hardware. Additionally, be aware of somewhat hidden costs, beyond salaries focused on this effort, testing the automated process, training cost and productive time lost to eventually become efficient at using the automation.

 

• What to do with exceptions to the automated process − What percentage of all the cases will follow the automated process and which cases should be handled as exceptions outside of the automated process? Are exceptions easy to identify? Can methods of handling exceptions be included in the automated process over time?

 

 

In my work with clients, after we have a clear strategy for growth, I frequently spend a fair bit of time focusing on processes and building models for processes. I ask this question: If business doubled, could you handle it the same way you are now? What generally results is that companies recognize that they have not paid serious attention to their current processes for a long time. The problem is that the processes in place have evolved without anyone having taken a step back to ask if this is the best way to do things. To really understand whether automation makes sense, you may want to start with a re-engineering effort (start with a clean sheet), as opposed to a process-improvement effort. Typically, performing process analysis with clients is quick, with some of the automation opportunities obvious and some processes too inconsistent (no hard and fast decision criteria that is a basis for automation) to automate.

Beyond that, determining the cost-benefit answer requires applying subjective judgment. There are no easy answers here, and even if automation is to be introduced in multiple steps, the order of its introduction is another question to be considered. Fearing their being inadequately prepared to grow, some companies automate too early or automate the wrong processes in the wrong order. More commonly, though, companies automate too late, and the expected economies of scale from growth are not realized because growth both uncovers and introduces inefficiencies.

One more thing: Companies readily automate customer-facing activities to save on the cost of an employee who answers the phone. While this may have obvious financial advantages, it can significantly hurt a company’s relationship with customers.

Automation – the how, what and when – is a key aspect of my consulting engagements related to growth. Let me know if I can help you.