Learning Curve: Making the Most of Outsourcing
Companies that view outsourcing as an easy way to offload commodity work are missing powerful improvements to be gained by working closely with service providers, says Professor Robert S. Huckman.
Companies that outsource merely to shuffle off commodity work to save costs might be missing important opportunities to work with vendors and significantly improve the final product.
"When it comes to the outsourcing of many complex professional services, the world is not as 'flat' as one might think," says Robert S. Huckman, the Albert J. Weatherhead III Professor of Business Administration at Harvard Business School. "In particular, when this type of work is outsourced it should not be viewed as a mere commodity."
Instead, he says, cultivating important person-to-person relationships with the vendor of outsourced services can improve the efficiency and perhaps the quality of services delivered—particularly in health care, where outsourcing is on the rise.
In research recently published in the journal Organization Science, Huckman and colleagues shine a light on how the outsourcing model fits the practice of teleradiology. The article, titled "Learning from Customers: Individual and Organizational Effects in Outsourced Radiological Services," was written by Huckman; Jonathan R. Clark (HBS PhDHP '10), Pennsylvania State University; and Bradley R. Staats (HBS MBA'02, DBA'09), University of North Carolina at Chapel Hill.
"When this type of work is outsourced it should not be viewed as a mere commodity."
Teleradiology is the reading and interpretation of CT, MRI, X-ray, and other diagnostic images. Increasingly, medical institutions outsource this work to licensed experts located across town or around the world.
The research team found that the more an outsourced radiologist works with a particular hospital, the more efficient he or she becomes reading that hospital's scans. In addition, that efficiency is not immediately translatable to other customers.
The outsourced radiologist
Some parts of the teleradiology model closely resemble outsourcing in industries like electronics, business processes, or software development. The health-care customer—typically a hospital—sees the need for a service that is reasonably well defined, but it can't provide the service at an efficient scale. Service providers see an opportunity to consolidate this volume, achieving the scale required to provide the service efficiently for a wide range of customers.
Here's how it works in teleradiology in the United States. Radiologists at an outsourcing service provider log in either remotely from their homes or at a reading center. As new scans come in, they are randomly assigned to the radiologists, who must be licensed in the state and credentialed at the hospital where the scan was performed. The random assignment of scans to any accredited and available radiologist suggests that the teleradiology industry views its product as a commodity—that is, two qualified radiologists are expected to read the same image in the same way.
That situation provided the research team an opportunity to test three questions about outsourcing that had received little attention before: Does a radiologist increase productivity by focusing on one customer? Does the radiologist's range of experiences with other customers result in better results delivered to that particular customer? Do individual's benefit from others' customer experience; in other words, does the outsourcing organization itself become more productive as its individual employees march down the learning curve?
"Though the increased use of outsourcing has led to a rise in the number of customer-supplier interactions, the learning benefits of customer experience remain largely unexamined," the researchers write in the paper.
The researchers studied data from OutsourceCo (a pseudonym), a major provider of outsourced teleradiology services in the United States, with more than 1,400 client sites, mainly hospitals and radiology group practices. The 97 radiologists in the study were all board certified and licensed to practice radiology in the United States.
The data set included 2.7 million scans read by the radiologists for 1,431 customers over a 30-month period. Nearly 85 percent of the scans were CT; roughly 10 percent were ultrasound; and X-ray, MRI, and nuclear medicine accounted for the remainder.
The scans were analyzed according to the reading radiologist, the customer (hospital or radiology group practice) it came from, the anatomical area depicted, and the length of time it took to read.
The results showed that experience between a radiologist and a customer paid off. A radiologist's experience reading scans from the same customer and involving the same anatomical area was found to produce the largest increase in efficiency, with an additional 1,000 cases of cumulative experience increasing a radiologist's subsequent weekly output by 7.4 percent.
Productivity is extremely important in this setting. Many scans are submitted by hospitals in emergency situations, so quicker analyses may lead to faster decision-making and better patient outcomes. The outsourcing firms, of course, also benefit by processing a greater number of scans.
The efficiency benefits of shorter read times did not appear to come at the expense of clinical quality; the rates of "discrepancies" reported by customers was low.
The researchers point to prior research to explain why. Repeated experience helps the service provider learn the customer's standard operating procedures, generally improves communication and coordination, and sets up the possibility of knowledge transfer between them.
The advancements accrued not just to individual radiologists, but to the outsourcing firm as well. Huckman explains what happens as an outsourcing firm gains experience with a customer. Let's say that the firm has 20 radiologists. Dr. A reads the first 200 scans from a particular hospital, developing considerable expertise and efficiency with that customer. Then she and Dr. B each read half of the next 200 scans from the same hospital. Both doctors have now developed individual expertise with the hospital—but so has the outsourced company as a whole. As other radiologists read more cases from the hospital, the difference made by the greater expertise of Dr. A declines.
As Huckman puts it, "Something gets baked into the firm's expertise that essentially substitutes for an individual radiologist having experience with that customer. If you're an outsourcing firm, that has implications for how you think about staffing a new client versus one with whom the firm has a longer history."
Huckman's future research will look at more instances of decentralization, including other types of telemedicine and specialized providers of niche services such as retail clinics, single-specialty hospitals, and disease management companies. The research, he says, reflects the traditional pendulum swing in health care between centralization and fragmentation of services. Some critics complain that health-care delivery is centralized in hospitals that are seen as too big and inefficient. In contrast, others find fault with an overly fragmented system that may result in significant coordination and quality problems.
Huckman says that his current findings speak to one of the key challenges posed by greater decentralization not only in health care but in other settings as well.
"Some services that we have thought of as commodities actually have subtle features that make them differentiated," he says. "Ultimately, the relationship between customers and the individual outsourced employees who work with them has an impact on the efficiency of the services provided."