2019-03-05 06:22:34    Robert E. Smith, PhD, Founder, The Change Shop

“Relentlessly Automating”: A Case Study on Decisions about At-Risk Jobs

Recently, PepsiCo announced plans to lay off a sizable (but as-yet-unknown) number of employees as part of a plan that involves, “relentlessly automating and merging the best of our optimized business models” in the words of CEO Ramon Laguarta. This was part of a 2-part road map Lagarta outlined for the second largest food and beverage business in the world. The first goal was to save $1billion annually through 2023 and the second, “…involves becoming more capable, leaner, more agile and less bureaucratic.”

“Relentless automation” isn’t just the latest corporate buzzword. Late last year, analyst firm Forrester predicted that automation could kill 10 percent of U.S. jobs in 2019, while creating the equivalent of 3 percent new jobs. In a similar vein, an October 2018 report by the World Economic Forum suggested nearly 50 percent of all companies expect their full-time workforce to shrink by 2022 as a result of automation, though, “almost 40 percent expect to extend their workforce generally and more than a quarter expect automation to create new roles in their enterprise.” While not a total loss, the macro-level picture seems to suggest workers are facing a new future of work where job loss will be a strong “maybe” and re-training to learn new skills will be a “must”.

The discussion about automation (human-replacing technology) made me a bit nostalgic about my doctoral dissertation where I examined the role job-threatening technology had on worker’s attitudes about changes in the workplace (I summarized the findings in, Selling Change: How Successful Change Leaders Use Impact, Influence, and Consistency to Transform Their Organizations). As part of my research, I cited a compelling study by MIT Professor, David Autor, that looked at how managers at Cabot Bank approached automation. This was the perfect case study for examining how managers think about jobs and ways jobs can be automated based on job tasks, people impacts, available technology, and costs. It provides a useful mental framework for thinking about factors leaders consider when considering automation.

Background

Cabot Bank is one of the top 20 largest banks in the United States. It has both retail and large commercial banking operations, with branches in several states and countries outside the U.S. At Cabot and many other banks, the job of “check processing” is divided into two main tasks. The first is deposit processing “non-exceptions” and the second is “exceptions-processing”. In the bank’s processing center, 2.8 million checks are deposited in the bank’s branches and ATMs daily. Until the mid-90s, a proof machine operator was responsible for manually sorting, processing, and ensuring accurate totals. In 1994, Cabot introduced “check imaging” technology that included high-speed cameras and Optical Character Recognition (OCR) software similar those that allow you to see the check you deposit in your local ATM seconds after you insert it (assuming, of course, you still use/receive checks). With the new technology, high-quality photos made the checks visible to anyone in the bank with access. Reading and recording check amounts no longer needed to be done manually by a person. The result was that the new technology eliminated the need for one of the key tasks performed by the proof machine operator job and divided the other tasks among more specialized jobs. Managers at the bank were responsible for figuring out how best to divide the work and this involved job and pay redesign.

2 Types of Management Decisions About Automation - Narrow vs. Broad

Managers had discretion to divide up the work in any way they wished but there was a financial (cost-savings) incentive for them to prioritize more specialized job tasks. Because automation reduced the need for low-skilled job tasks (e.g., removing staples and ensuring the checks faced the same direction), managers were incentivized to make the remaining jobs more specialized and higher-pay. Because automation technology reduced the need for some of the more menial job tasks, managers on the downstairs floor of the bank opted to make the check exception-handling job more specialized. In the case of Cabot during this period, there were no lay-offs as the bank was experiencing growth through acquisitions, but productivity did increase significantly (27%) nearly tripling the amount of checks that could be processed nightly with the workers they already had onboard in a 10-year period. I call this technology-focused job redesign ‘narrow’ because the changes were done at the job and task level.

“Due to improvements in OCR software, a growing fraction of checks are read without human intervention. One result is that the demand for keyers at the bank per million checks processed has declined.”

On the upstairs floor of the same bank, the Vice President in charge of exceptions processing had an even grander vision. Using language similar to PepsiCo’s CEO Laguarta and CFO Hugh Johnston, the VP at Cabot believed that a broader reorganization could achieve three goals: 1) improved productivity, 2) better customer service, and 3) better jobs using more skills. To quote the Cabot VP, “fewer people doing more work in more interesting jobs.” Let’s call this type of job redesign ‘broad’.

Rather than dividing the check-processing work on the basis exception vs. non-exceptions, this broad redesign effort aimed to divide the work on the basis of customer segments so that this new Customer Service Representative (CSR) role would have a much bigger scope such as verifying signatures, handling overdrafts, and coordinating or communicating exceptions. They would be expected to do more and know more across the full scope of services the bank offered to their customers. Beyond the sweeping rhetoric, the bank leader in charge of this initiative also employed some of the key principles of effective organizational change to move employees in the new direction. He ensured current employees were involved in the job redesign and conducted focus groups where managers would ask the check processors what aspects of their jobs were irritating and what changes would make their jobs better. The VP at Cabot believed that motivated employees, automation, and job redesign all had a role to play in making the bank more productive. He said, “If you do your job, you get to keep your job—but you may not get cost-of-living wage increases. If you transform your job in a positive way, you will get a raise. If you transform your job and have a positive impact on the people around you, you will get a promotion.”

Bringing it All Together

Cabot was one of the first banks to reorganize exceptions processing using check imaging and therefore provides a model for how organizational leaders approach automation. Between the downstairs and upstairs floors of the same bank, managers at Cabot took radically different approaches to technology that had the power to eliminate jobs. In the downstairs floor they took a narrow, task-focused approach that involved little input from workers and resulted in a 27% increase in productivity (e.g., number of checks processed). In this narrow approach, really good check exception handlers were moved into higher-wage jobs that paid 16% higher wages than the previous jobs that were automated but there were no lay-offs (although there easily could have been had the bank not been in a growth mode).

Alternatively, on the upstairs floor, the bank leader took a broad, job-redesign focused approach that grouped the work of all exception-handling tasks by customer segment (rather than task). Although a more expensive undertaking (e.g., upfront training costs and additional time), this broad redesign resulted in a similar, nearly 28% increase in productivity with the number of employees needed to handle exceptions decreasing from 670 to 530 employees with job redesign-only changes and again from 530 to 470 employees after the automation technology was introduced. It also resulted in an approximately 21% increase in pay for the employees who were able to retain their positions.

“…the business case in favor of reorganization became compelling only when managers knew that the productivity gains from reorganization would be enhanced by time savings from eliminating paper shuffling that image processing [job automation] made possible.”

The Results: Key Themes in Job Automation

In the above case of the two floors / and two differing managerial approaches, automation made it possible for bank leaders to ‘think differently’ about what could be done with jobs and work tasks. In both cases, automation eliminated the need for low-level job tasks while increasing the emphasis on higher-level job tasks requiring more skills. This led to an increase in the need for more training, more skills for more ‘complex’ work, and higher pay for those who remained who could perform the higher-level job roles. Finally, fewer jobs were needed to produce significantly more output. This case study sheds light on common ways job-threatening technology can be introduced and what it means for workers, productivity, and job design.

The Great Technological Unemployment Debate

Technological optimists do not believe the rise of cognitive technology, or machines that can approximate the way people think, will lead to massive unemployment. IBM CEO, Virginia Rometty, insists that automation, “…is not about replacing people. It is about augmenting what man does…this helps us do things we couldn’t do.” IBM is famous for its Watson cognitive computing system that beat two Jeopardy champions in 2011. Others take a more pessimistic view of technology’s impact on workers including MIT Professor, Andrew McAfee, and co-author of The Second Machine Age, where he argues that productivity growth has been decoupled from jobs and income and in the digital economy, goods and services can be provided to an infinite number of additional customers, all at the same time, at a cost that is often close to zero. The case of Cabot bank seems to provide practical evidence for both views. Automation made it possible for jobs to be completely reconceptualized (broad job redesign) while also making it possible to eliminate entire job categories (narrow task redesign).

What’s your experience been with job automation?

Case Study Source: David H. Autor, Frank Levy, and Richard J. Murnane. Industrial and Labor Relations Review, Vol. 55, No. 3 (April 2002). © by Cornell University. 0019-7939/00/5503