The Wavelength: Automation and Innovation
Automation: It’s a big part of innovating today’s workflow. We asked a panel of experts some questions about the ways in which automation and innovation are critical to today’s businesses.
How does data analytics play a role in process automation?
Thomas Dong: Analytics fuels automation by enabling more intelligent and informed action at key decision points in any digital process. By leveraging data, including both structured and unstructured, as well as historical and real-time, analytics can help synthesize endless amounts of data for broader and more complete context, pinpoint patterns and trends, and then recommend the next best step whenever a decision point is triggered by rules. For instance, analytics-based process automation can help industries such as manufacturing predict supply chain disruptions and recommend alternative routes or suppliers, optimizing the process before the disruptions occur and minimizing any impact to SLAs. Or think about the analysis of social media channels of a business, where analytics can be used to not only identify a triggering event, such as a threshold of negative comments, but then respond to the triggering event with a highly personalized marketing campaign (e.g., right offer at the right time). Analytics has a huge role to play in modern process automation and will help organizations better prioritize limited resources, act more quickly, and avoid or exploit likely outcomes, all resulting in higher performance and efficiency while lowering risk.
Ed McQuiston: With the evolution of both the products and the definition of process automation over the years, data analytics has taken on a much more critical role, and one that has also evolved. A big reason why is the kinds of information being managed by process automation tools. Long gone are the days of just document-centric processing. … For many years, “data analytics” around process automation was defined and focused on retrospective productivity measurement: “How much more efficient is my process?” “What are my turnaround times?” “Am I meeting my process time SLAs?” These questions were answered by reports run post-process to measure processing times. Later, those analytics became all about “real time” monitoring — measuring and monitoring work in progress (WIP) to enable load balancing and efficiency gains by evaluating bottlenecks, peak periods, seasonality and more. Today of course, analytics has become as much or more about business insights as it is productivity. With either standalone or embedded business process management tools often acting as a middleware between disparate core systems, data analytics of the information flowing through these process automation tools can often be the place to gather an aggregate view of typically siloed data, increasing insights across a customer relationship, a patient lifecycle, an insurance claim or a student at a university by example.
Marie Rosecrans: Data analytics can show inefficiencies within a business, which automation can then resolve. For example, analytics can give a company insight into the average time it takes for their service reps to close cases. If service standards are not up to par, companies should evaluate ways to automate processes and speed up performance — like automatically answering frequently asked questions so service reps can focus on high-priority cases that require extra care.
Jupp Stoeptie: In today’s business environment where customers have expectations of “now” for services and want a seamless relationship with their vendors, governments and suppliers, analytics is proving to be an effective tool in understanding customers, their behaviors/preferences, and to improve the customer experience overall. Analytics not only ensure the fastest path to resolution or purchase, but also the next best step in a process while making data visible and understandable for more effective decisions. Organizations investing in analytics are achieving competitive advantage.
How is automation changing the business environment?
Joe Odore: For years, automated equipment and processes have revolutionized almost everything business does. Automation touches every task from scheduling deliveries, to sending out invoices, to extracting data for annual reports, to relegating inboxes from desktops to trashcans via secure email. Cost savings are and will continue to be the ultimate driver in business automation. For example, in transportation today, trucks may be automatically redirected to optimize cargoes and maximize sudden tariff opportunities that might well go to a rival carrier in the time it would take for freight managers to debate and decide on a redirect strategy. Future automated data gathering, analysis and decision-making programs will be enhancing bottom lines by instantly rerouting shipments from rail to truck.
Rosecrans: From sales, marketing, customer service and more, automation enables employees across every department to free up time and focus more on building stronger customer relationships and better experiences. For example, most salespeople spend 34 percent of their time on one-to-one interaction selling, according to CSO Insights; the rest is dedicated to other operational tasks. A CRM system with automation can streamline processes like assigning leads, updating opportunities, creating quotes, gathering signatures and more. Automating these repetitive tasks frees up more time for salespeople to focus on what’s a priority for them: building strong customer relationships and closing larger deals.
Stoeptie: Automation is drastically impacting the competitive landscape of many businesses. Just think about the last time you walked into a bank, used a phone to order a taxi, or waited in line for a paper ticket at the airport. Automation is all around us and is driven by demands of customers. Organizations stuck in processes dependent on human capital are at risk from their agile, digitally focused competitors. According to KPMG, “Successful digital labor automation projects typically see cost takeout in the range of 40 percent to 75 percent.” This means that enterprises prioritizing digital transformation are not only faster and smarter, but also more efficient. The impact of the focus of automating digital labor is not only better response times and more effective processes, but the reduction of human capital focused on tasks that are solved by robots or automation.
What are some of the most innovative ways mobile technology has been integrated into automation?
McQuiston: Well, I think the industry has just scratched the surface here. But whether we are talking about employees in the field, customer interactions or just the reality that many of us are working on form factors other than desktops or laptops, we are all constantly interacting with our phones and tablet devices. On-the-go executives provide reviews, approvals and even sign documents on their phones. Service employees on remote oil rigs review plans in order to perform services. Construction and development companies exchange designs with engineers on active sites. Literally thousands of applications exist where mobile technology extends process automation to previously untapped solution areas. Beyond just displaying or interacting with centralized information, leveraging the mobile device itself means being able to leverage things like geolocation as a core part of any process automation.
Rosecrans: Businesses today can tie all customer service experiences directly into their mobile apps. This is made possible with the availability of mobile software development kits (SDKs), which consists of tools to build apps with customized user experiences. Using a mobile SDK, companies can build a virtual help desk integrated into their company’s app and embed their knowledge base, allowing customers to benefit from any service-related automations currently in place. For example, when a customer fills out a service request, the mobile SDK can automatically search the knowledge base for relevant articles and offer them to the user right from their phone.
Stoeptie: Digital enterprises are adopting mobile-first initiatives and benefiting greatly. Customer reliance on the mobile device as their gateway to vendors, suppliers, and solutions means that new account opening, customer onboarding, claims, lending, purchasing, payment, and proof of identity solutions must be digitally born, multi-channel and mobile first in order to keep pace with customer expectations and market demands. Real-time recognition capabilities that turn the content within the viewer of your mobile phone into hands-free processed data is dramatically changing the user experience and customer experience standards for mobile apps. Whether automatically capturing ID information for a new account, translating a menu, or reading a license plate or VIN to process a claim, mobile apps that incorporate real-time recognition are improving usability, acceptability and improve the overall customer experience. The most common mobile use cases we see involve some level of proof of identity and the process of capturing supporting or trailing documents driven by a lending, mortgage or claims process.
What advice can you share for identifying specific processes that could benefit from automation?
Dong: Traditionally, the bias has been to only look for “big” processes to automate, but we are seeing companies gain significant efficiencies by tackling several smaller point processes. With that in mind, begin by looking across the organization and identify any workflows which can be simplified. Good places to start are with manual processes that often connect business processes from one application to another. Excellent examples of these are traditional case management processes around incident reporting, customer inquiries and various types of approvals. Additionally, don’t overlook business processes like the creation of marketing assets or the exchange of digital assets with vendors, which can be automated to provide greater visibility as well as provide often needed audit trails. Once the process has been identified, consider the associated tasks and determine which ones are repetitive, thus requiring simple automation, or those that may be more complex and would benefit from data-driven automation. Many of these processes cross systems, which may have been excluded in the past but can be easily addressed today. Your goal should be to find areas that are open to efficiencies to help enable a “working smarter” approach, since automation helps open up skill sets to focus on more complex tasks and refocus the organization on its core business.
McQuiston: I think it comes back to changing the mindset as to what process automation is, bearing in mind that processes can be driven by any kind of content. Discrete data from line-of-business systems, web forms, scanned-in documents, office content, faxes, PDFs, even a digital photo, audio file or video could be part of a process that would benefit from improved automation. And additionally remembering that process efficiency is only one reason to look at process automation opens up a different lens as to what could benefit.
Stoeptie: I often tell customers that our goal is to automate cognitive tasks. Any time a user or human must look at a document, find information within the document, or enter information from a customer or document, we want to help them automate that process. The advancements with AI, machine learning and natural language processing are making this automation a reality and customers are not only reducing costs, but dramatically improving process outcomes for customers and businesses alike.
Are there processes that can’t be automated?
McQuiston: While I’m sure there are, it’s important to return to the definition of automation. More and more, customers are asking for help in automating processes that are not serial in nature. In fact, the data or document is no longer the heart of the process they are looking to automate. Rather, it is a series of tasks – a checklist if you will – that must be performed. The content is really secondary to the completion of the tasks. Case management has proved to be a different type of process automation for customers but one that is incredibly helpful in addressing processes that were previously not well suited to older process automation tools.
Stoeptie: Not yet, but there are certainly cost concerns related to training a system, gathering samples, evaluating results and then fine tuning. Some processes are not being automated simply because of the investment in resources to train an automated system may not deliver the immediate ROI and recognition results. But I can tell you that advancements in AI are greatly reducing these cases.
How specifically can automation help the SMB innovate?
McQuiston: Ironically, this is a severely underserved market, and one that often has huge needs to automate. In order to avoid the typical costs of growing a business, inefficiencies are often accepted in order to avoid the costs associated with people/process/technology. A few years ago, I worked with a customer that was on the cutting edge of technology. There were hundreds, if not thousands of developers. The absolute latest technology on everyone’s desk. These guys were state of the art, but a very young company — less than 10 years old, but they had exploded. What struck me was that their processes were very manual. They were shipping contracts all over the world for review (10 copies to 10 different places). They were overrun with file cabinets. They were literally operating as a company might in 1975 on the process side of things, but this was a cutting edge technology company. How could this be? It turned out that when they were small – tiny in fact – they never took the time to automate. They focused only on growth. Then they got too big to slow down and course correct on their processes. Unfortunately, then they got too slow to keep up with demand because their processes didn’t support the size and scope they became. In this day and age of cloud-based solutions, subscription pricing models and packaged content services with specific process automation tools, SMBs have better opportunities than ever to take advantage of process automation opportunities earlier in their maturity cycle.
Rosecrans: SMBs have limited time and resources. Automation lets them spend fewer hours on tedious operational processes, and more time innovating and setting their businesses up for success. The challenge we’re seeing is that many SMBs still manage their businesses using manual tools that can lead to data siloes and limit the opportunity for automation. For example, a survey we conducted found that 44 percent of small business owners use email to track their customer information and 41 percent use spreadsheets. When they automate, SMBs enable employees to focus on larger objectives and growing the business. For example, a CRM system can store valuable information such as customer contacts, purchase history and past interactions. From there, the technology works its magic and uses the data to automate tasks such as identifying customer cases that need immediate attention, surfacing priority leads, or even something as simple as entering customer information. If automation is implemented across the organization, it allows employees to focus more on the creative aspects of their job and bringing innovations to the business.
When automation is introduced to a process, where are bottlenecks most likely to occur?
Dong: Traditional process automation focused on reducing risk and variation in straightforward, repetitive processes. It then evolved to case-based processes that engaged the knowledge worker to drive the process through critical steps. Now with data-driven automation, complex case-based processes can be automated, further reducing the need for knowledge worker involvement. But with AI or machine learning now being leveraged for “next best step,” a new set of bottlenecks around data access has emerged, as well as new procedures for engaging knowledge workers on exception handling. Meanwhile early apprehension toward these technologies have also created some additional bottlenecks. However, as knowledge workers gain trust in these systems, the benefits of automation will be the ability to focus on higher value activities.
Stoeptie: It is always fun to work with customers and see where automation delivers some very important benefits, but also exposes additional challenges that were not foreseen or projected. We see manual process approval queues fill up now because transactions are moving straight through from receipt, so the historic steady stream of transactions moving through the workflow in a day are now processed in real time and delivered to a human for final review, approval or escalation. This visibility and transparency of the new challenge is always a welcome sign because it means that we delivered on the upfront automation benefits and companies normally have human capital available to be allocated to the new bottleneck.
Should people be afraid of losing their jobs to robots?
McQuiston: Ha ha. Well, I think I’d be more afraid of getting run over by self-driving cars on the way to work than losing jobs to robots when it comes to business process automation. Transparently, there is no denying that the enhanced efficiencies brought by technology have changed the employment landscape all over the world. “Low value” tasks have been targeted for automation by companies all over the world and by and large technology has been used to do that. However, the role of the knowledge worker really remains unchanged. They may sit in front of a different system today than they did, but decisions based on business understanding still must be made. A robot does not know empathy or compassion. A robot doesn’t understand business or customer impact that goes beyond the parameters it is programmed to understand. While I would be the first to submit that we are nowhere near done automating new processes and potentially further changing not only how people interact within organizations, but how many, there will always be a place for people who understand a process, its impact, ways it can be improved and its effect on the customer. And there will be continuous improvements to the technologies those people use to do their work.
Odore: Yes and no. How many bank clerks and supermarket cashiers have been “let go” in favor of self-service checkout stands due to the ATM card? ATMs, for all intents and purposes, are robots. They perform every phase of a standard workflow process, from accepting payments and making changes to ordering you to bag your purchase at the supermarket checkout counters all with robotic efficiency. They’ve contributed to the downsizing of thousands of bank teller and cashier jobs. Would those people who’ve long since had to find new jobs trade things for giving up their ATM debit cards and going back to waiting on the queue? Robots don’t have to be used to replace people to increase a corporation’s profit, they can also be used to persuade people to work for lower wages. For example, all the banner headlines about driverless cars and the workforce have focused on Uber and Lyft. In 2012, when Uber rostered virtually zero full-time U.S. drivers, the Federal Motor Carrier Safety Administration reported that nearly 6 million Americans were earning (or trying to earn) a living by driving a truck. Technologically, the R&D resources devoted to robotic trucks have met or exceeded those being expended on robot cars. The real question is will the major trucking companies be willing to pay the token “drivers” sitting in the cab watching DVDs the same $70,000-plus a year many are earning today?
Stoeptie: Robotics and AI are the inevitable next wave of automation that will drive our global economy for growth. Naturally jobs will be impacted, no different than when transformational technologies impacted manufacturing, transportation, communication and agricultural industries. The important aspect of this transformation is our ability to reallocate and retrain impacted personnel to deliver higher value tasks for the enterprise and let the robots handle the repeatable tasks as capable digital workers and assistants.
This article originally appeared in the September 2017 issue of Workflow.
