Taking the Robot out of the Human
Robotic process automation (RPA) is steadily gaining steam in offices around the globe.
This software, which is designed to carry out simpler, more repetitive human tasks, such as data entry, accounted for about 7% of repetitive work in 2014. It will rise to 23% by 2020, according to a Willis Towers Watson study. That means that, by next year, RPA will be doing nearly a quarter of back-office chores.
As Leslie Willcocks, a professor at the London School of Economics who studied technology, has said, the real value of a robot is that “it takes the robot out of the human.”
How can brokers take advantage of automation? By proactively transitioning employees whose jobs may be changing and by seeking new ways to serve their clients through a combination of traditional skills and the latest technology.
Robotic process automation can take a high volume of data and transfer it from one software system to another. In other words, it can do such things as read a form at a carrier and put the data into a spreadsheet at the brokerage office. From there, the brokerage can use algorithms to analyze the data for a client. RPA also reduces human error in transferring data, one of the biggest obstacles to gathering accurate information in the insurance industry.
“Anything we do on a routine basis today…is ripe for automation,” says Brian Comerford, senior vice president and IT director of strategy, governance and innovation at IMA Financial Group.
He cites an example: let’s say your job requires you to spend a few hours each week contacting carriers and retrieving information to provide a quote for a client. RPA can do this for you. First it would visit selected carrier websites, then pull data from their forms, then transfer it to your own spreadsheet at the office, and finally deliver it back to you in whatever shape you’d like it. Time saved: seven to 10 minutes per transaction. Now multiply that by a dozen or more times per week. Suddenly, you’re saving an hour or more of your time—by the end of the month, maybe even an entire day or two. That quickly becomes real money.
And that’s only one labor-saving example. Robotic process automation is evolving in the office as employees discover individual software bots they find efficient to work with. “A slew of them are coming out, and workers will find what works best,” says Brian Kane, owner of General Design and a specialist in artificial intelligence and user experience.
Robots can do tasks you might not think possible. Already, a writing software called Quill, developed by Northwestern University and marketed by Chicago-based Narrative Science, can actually write earnings statements for Forbes magazine and business reports for T. Rowe Price and other clients in financial services.
Another way to describe it: there goes the back-office staff. Such a projection might be an overreach at the moment, but it appears to be the future. When fully implemented, RPA could eliminate the need for humans in the back office. In Japan, for example, one insurer has already laid off workers in favor of robots. Fukoku Mutual Life Insurance laid off more than 30 workers two years ago and replaced them with RPA.
To a great extent, industry observers predict that underwriting, pricing and claims will be taken over by robots in the next decade. According to Oxford University, out of 702 professions studied, the fifth most vulnerable job of all professions is insurance underwriter. You read that correctly. They are barely more sustainable than telemarketers, who were named the most vulnerable profession in the study (which comes as no surprise to anyone who has received those irritating dinnertime robo calls).
“A great deal of underwriting will be automated in a matter of seconds,” says John Boudreau, professor and research director at USC’s Marshall School of Business and Center for Effective Organizations.
The 14th most vulnerable profession projected to be replaced by new technology is insurance claims and policy processing clerk. Insurance auto appraisers are ranked 18th most vulnerable. Claims adjusters, examiners and investigators rank 27th.
And insurance agents? They ranked 137th most vulnerable of the 702 professions studied—in the top fifth of most-vulnerable jobs. Even utility meter readers are considered more irreplaceable than agents, ranking 217th on the list.
One forecast from PwC estimates nearly a third of all jobs in the financial sector (including banking and finance) will be taken over by robots in the next 20 years. And insurance isn’t the only profession that will be hit hard. The Oxford study predicts computer algorithms will replace nearly half of all lawyers and managers in the next two decades.
Re-engineering Your Job
The possibility of changing the entire commercial brokerage landscape is huge. Some jobs may be eliminated, but most will be re-engineered into new jobs, with different responsibilities working side by side with automation—so-called cobots, or collaborative robots.
“The job of the broker will indeed change,” Boudreau says. “But that job is likely to interact more and give up parts to other places in the supply chain.”
Brokers now have a computational role when putting together proposals for clients. Their involvement in pricing and underwriting is likely to decline as technology expands and takes over that role. Instead, their sales and interpersonal communications skills, which brokers actively flaunt as their competitive advantage today, will become even more highly prized in the future.
“The body of work of this individual will shift from what they are doing to that part of the skill set where there is a true return for creating an enduring relationship with the client instead of manipulation of the numbers,” says Ravin Jesuthasan, managing director and global practice leader at Willis Towers Watson Talent and Rewards.
“Automation removes the bottom part of the old job,” Boudreau says. Brokers might not be hired for their expertise in risk and pricing anymore. Instead, their value will be based even more on relationships but with a twist: brokers will be forced to become more savvy in their use of technology, especially artificial intelligence.
“The best people in these jobs will be the ones who work with AI and embrace it,” he says. “Automation elevates the person who interacts with the customer to become the superhuman of the workforce.”
One reason for this is simple, Jesuthasan says. Clients’ needs vary, and so do their broker relationships. With artificial intelligence at their fingertips, brokers can delve into a range of areas that affect clients’ ability to run a successful business. “At that instance,” Jesuthasan says, “the broker becomes the pivot point because the carriers will be providing a commodity.”
Ben Fidlow, global head of core analytics at Willis Towers Watson, sat down with Leader’s Edge at RIMS 2019 and shared some of the ways his brokerage is doing just that. Willis Towers Watson made a decision to move away from customized analysis in tools like Excel and instead built a platform for scalable predictive modeling. The platform, called Connected Risk Intelligence, takes all of the previously compartmentalized analyses and aggregates them to create an overarching risk lens.
“It’s easy for us to do and easy for us to apply,” Fidlow says. “We can do predictive analytics almost on the fly.” The platform also has a presentation layer, which Fidlow calls the consumable part.
“Gone are the days of the 80-page PowerPoint. Now it’s dynamic and interactive, and we can pull in other things to help us evaluate risk,” he says, like a client’s EBITDA or cash flow. “So we can tell a better story, and I think that’s going to change even more as we go forward.”
And this is where it becomes a true differentiator. Fidlow explains that the ability to plug in different data and risk projections allows Willis Towers Watson to engage not only in more risks but also in other business areas. For example, he says, “Here’s your 10-year financial projection of your firm. Fine, let’s plug and play a risk analysis in here. OK, if we can reduce your risk this much, well, that frees up capital. What do you want to do with that capital?
“I think the world is really changing. I think our industry has been slow to get to it, but we have a couple of advantages: we’re good at data; we’re good at risk analyses; and we can empower our corporate clients who are trying to figure out what to do with big data and analytics.”
That means branching into business process areas such as manufacturing process optimization, sales and M&A activity. Willis Towers Watson has already begun delving into its clients’ loyalty and rewards programs using their technology.
“We’re used to dealing a lot with data, and these corporations are trying to figure out how to improve their business processes,” Fidlow says, “but they’re not focused on the predictive modeling side, so I think there’s a space for us to really dive in.”
It’s a Culture Thing
While the opportunities are there, redefining a business strategy based on technology is a process. Five generations of workers, Jesuthasan says, have toiled under a single, unwavering ethos: “I learn. I do. I retire.” Now he says, career paths will change. “Your ability to stay relevant is to constantly learn and reinvent yourself,” he says. “That’s a big ask of society.”
Change comes in fits and starts. As an example, Boudreau says, we upgrade our cell phones almost annually, yet we don’t need to relearn a whole new system of using it until the manufacturer makes some radical change. Careers will be similar. While change is constant, it is also incremental and not necessarily disastrous to your position in the firm if handled with insight and planning. “It is a business culture thing,” Boudreau says. “This needs to be a continuous conversation, and it needs to be at the task level, not the job level.”
Jobs will be re-examined, not because an employee’s role will disappear but because it might be altered with new responsibilities and tasks dictated by new technology. Businesses are capable of seeing the evolution in the office in advance, Boudreau says, so they should plan transitions early and avoid staffing issues.
“Are we going to wait until it disrupts you?” he says. “How do we humanely move someone out of their job? Or do we start it earlier? Let’s have the courage to talk about it earlier. Let’s not start talking about people losing their jobs. That’s a nonstarter.”
Note: Jesuthasan and Boudreau co-authored a book, Reinventing Jobs, A 4-Step Approach for Applying Automation to Work (Harvard Business Review Press, 2018), which examines the effect of the rapidly emerging revolution in office automation. Their book explores how such automation will help create optimal human-machine combinations and how future business careers will be constantly reinvented as a result.
Collaborating with Robots
Alexa is just the start. Here’s what could be coming to
Voice-activated computing using natural-language processing—the ability of a computer to understand, interpret and act upon conversational speech—is currently dominated by Amazon’s Alexa. Alexa is mainly known as a personal assistant in more than 20 million homes, not for its office prowess. People talk to Alexa, and it talks back, giving them the weather, turning on their lights or television, playing their musical requests, reading the news and much more.
But Amazon wants to get into the office environment and currently has 5,000 employees working on Alexa for home and office. Amazon’s Alexa for Business involves placing Alexa devices all around the office to handle any number of office staff tasks. Eventually, it could ease conference call malfunctions, operate office machines, order supplies, enhance communication within the office and conduct any number of office tasks—as soon as someone can imagine what they might be.
Imagine sitting in a meeting and immediately ordering up sales figures on a wall monitor or a report on some new division or product simply by asking Alexa, which is sitting on the conference room table ready to go to work.
Cortana, Microsoft’s version of a natural-language processing digital assistant, has the advantage of being compatible with Microsoft Windows and Office software as well as Android phones. That gives it a huge advantage in the office technology competition. Google and Apple have also entered the field with their own devices. Whose natural-language processing device will dominate the future work environment? It has yet to be determined.
Natural-language processing gives you the ability to synthesize and connect knowledge across a large organization, says Ravin Jesuthasan, managing director and global practice leader at Willis Towers Watson Talent and Rewards. Using artificial intelligence and machine learning, it will eventually prompt you with questions to help refine and narrow your searches even faster.
But there are problems with this, says Brian Kane, owner of General Design and a specialist in artificial intelligence and user experience. “Spoken interfaces have some human-factor issues that affect workplaces, essentially the noise and privacy issues involved with people talking at work,” he says. “But these services have text interfaces, so I see most of the value there.”
Kane does not foresee spoken interfaces replacing screens in the office. “That’s why they are starting to add screens to voice devices,” he says. “[Spoken interface] isn’t always better or faster.”
Imagine, Kane says, if everyone in the current open-office environment were talking to a natural-language device. The noise could get unbearable, lack of privacy would escalate, and the work environment would deteriorate. Even without this escalating auditory distraction, Kane notes, we already have employees wearing earbuds or headphones as they work to block out the noise in the office.
Could natural-language devices change the work environment back to individual offices? Technology has its ramifications.
“I’ve done some research on human interfaces to robots, and I think it’s a lot more complex, from a human-factors standpoint, than many people realize,” Kane says. “Specifically, I’m interested in how people will communicate with robots in a work or service environment. When you have a controlled one-to-one relationship between a robot—physical or digital—and a person, it’s much simpler to have good communications. But when you start to have multiple people and multiple robots in an unmanaged or semi-managed environment, the complexity of the interactions grows exponentially. We’ve all seen sci-fi movies where people talk to robots, but in the real world, I’ve come to think that a better way to communicate with robots is through our smart phones, primarily with text. I think we will ‘friend’ robots on our phones and then have robot friends in our contacts that we communicate with via email, text, voice and video.”
Other Office Mates
Machine learning is used for chatbots to answer online questions from clients or to process an inquiry before it is handed off to a staff person. As questions are repeated, the software “learns” a proper response that it can later use to narrow the scope of an inquiry. This allows it to engage a client longer before handing the client to a staffer.
Collaborating robots, known as cobots, work side by side with human cubicle dwellers. These robots use advanced software bots built with machine-learning algorithms. They currently exist in such businesses as Walmart and Ernst & Young. Employees work hand in hand with them and share tasks to get the job done.
As for stand-alone robots, there are some burgeoning uses. A robotic receptionist can greet visitors and notify employees they have a guest. Japan has even developed a robot that can read the news like a television news anchor.
A telepresence robot can roam the halls of the office, allowing an employee thousands of miles away to access meetings and engage employees via a mobile camera and microphone. The offsite employee simply calls in to the robot to participate in office life. The robot looks like an iPad on a stick atop a Segway, which allows office workers to see the offsite employee’s face on the screen and talk with him or her. Other robots roam the office at night, acting as security guards.
So what will robots make the office of the future look like? Kane believes it will look much like it does today. “But with more automation of routine tasks for document processing,” he says. “Translation services are going to get much better, and that can really help with international business. I think making and posting web content will get much easier. You may see individual workers find the best tools and applications on their own, and then they can be brought in-house after they have shown their value.