James Thew - Fotolia
Software robots will increasingly be deployed in the service industry. The financial services sector is already using digital labor, as it's sometimes called, to automate routine, rules-based jobs traditionally handled by humans. And other industries are expected to follow suit. CIOs who are not exploring the use of digital labor are depriving their companies out of significant efficiencies and process improvements.
That was the take-home message from Stanton Jones, director of research and principal analyst with IT advisory firm Information Services Group Insights, at the recent Digital Business Summit in Chicago.
Jones defined digital labor as software that performs work like a human or mimics the way humans make decisions. In this SearchCIO Q&A, he explains the different types of digital labor, potential use cases, risks and benefits, and offers advice on how forward-looking CIOs can get started on digital labor.
What are the different types of digital labor and potential use cases?
Stanton Jones: There are three levels of digital labor. I would like to use an analogy to explain this. Let's say you have a map and you are trying to get from point A to point B. You know only one route on this map, and you want someone to follow that route every single time: That is something I would call Level 1 of digital labor. The application of that is through a technology called robotic process automation. Those RPA systems automate and work like a human, and they are mainly being applied in areas like HR, finance and accounting, and customer care. Level 1 is automation of 'swivel-chair' kind of activity [in which humans must navigate between multiple applications to perform a task].
Level 2 is when you have the map and you have several possible routes to your destination; this type of digital labor helps you find the best route to your destination, depending on the conditions of wherever you are travelling on this map. I would call these expert systems, and these are primarily being applied in areas like IT. This Level 2 digital labor we see in vertical use cases, like in banks and insurance companies. They are using these expert systems in fraud detection and compliance.
The last level would be Level 3; this is where you have the map, but you don't know how to get to the location you are trying to get to. These would be machine learning systems. So, these kinds of systems learn based on giving them a lot of data; in my map analogy, it will create the route for you, depending on what it learns from the data.
How is digital labor transforming the enterprise?
Jones: At this point, we see real adoption of Level 1 and Level 2 digital labor within the enterprise. These systems are making existing employees more productive; employees are getting more work done than they had in the past. In longer term, that's going to have implications, meaning if you are being more productive, you don't need to hire as many people in the future. But, right now, we don't see this as taking jobs. In several use cases, it is actually improving morale internally because it is taking a lot of the lower-level work, the repetitive work, and letting people focus on more interesting work.
At the machine learning level, there is a lot of experimentation happening, but not much real implementation yet. It's partly because the technology is very new and there is a massive shortage of resources to build this kind of technology.
What are the benefits and risks of digital labor?
Jones: Work is getting done much faster than it did in the past, and it's also getting done with more consistency; there is also better quality of work because the expert system, or the RPA system, is applying this set of rules on the workload the same way every time, and it doesn't make mistakes like humans do. Ultimately, that is going to lead to cost reduction because you are going to be able to do more work with less people. A benefit of using digital labor will be future cost avoidance.
The software platforms that create digital labor are new, and most of them are very small. For most enterprises that are jumping into this, that would be one of the risks. Enterprises are putting a lot of time and effort and energy into essentially virtualizing some of their human workforce ... they have to understand that there will be a consolidation wave at some point, meaning the big companies will start acquiring these smaller platforms. Companies have to make the right bet in terms of digital labor platforms they invest in.
I think digital labor is going to disrupt a lot of labor-intensive providers in some ways. Big outsourcing providers have a huge number of people in terms of their labor force. Digital labor turns out to be very disruptive for them because it potentially virtualizes their labor pool and they have for many years focused on how many people they can get working at a client because that equals to revenue.
What role do CIOs play when it comes to introducing digital labor in their organization?
Jones: I am of the opinion that CIOs should take the lead because they are in a prime position in their organization to drive digital labor into the enterprise. There is nobody in the enterprise [who] understands better the virtualization of a physical thing; CIOs have been doing that for years with virtualizing servers.
CIOs also, by nature of their role, are right in the middle of all of the processes that happen within an organization because all these processes require technology, so CIOs understand the implications of digitizing something physical, and they also are in the best position to understand how these processes are linked and interlinked.
What's your advice for CIOs who want to start deploying digital labor?
Jones: First, [CIOs need to] identify an automation or digital labor champion. It maybe the CIO himself, or maybe one of his direct reports. Or it may be somebody outside of his organization that he can encourage to move this forward. This 'champion' may not be a technologist, but it must be someone who understands the power of the technology and can drive this organizational change and who understands this from a process and operating model perspective.
Second, create a digital labor center of excellence composed of a small number of technologists, process people and analysts who can actually take these use cases and analyze them, redefine processes, apply the digital labor to the processes and then track results.
Third, use Agile approaches ... identify an area where digital labor can be applied, design the process, build the automation and get it into production. Using this kind of fail-fast, minimally viable product approach to this is critical, because if you wait too long, it is going to pass you by.
Read about the emergence of RPA software.
Read why machine learning algorithms are dominating software.