Too-frequent feedback can sink workers’ performance

New research by Assistant Professor of Accounting Pablo Casas-Arce finds professionals do better when they receive detailed assessments less often.

From C-suite executives down to students on their first jobs, most workers receive appraisals and feedback on their job performance.

If that feedback helps workers learn and improve, the traditional thinking goes, then more frequent and more detailed information ought to lead to even better performance. But new research finds that constantly opening the information floodgates makes people more likely to sink than swim.

The research by Pablo Casas-Arce, assistant professor of accounting, and colleagues at the Universidade de Lisboa in Portugal and the University of Notre Dame found that workers perform best when they receive detailed feedback less frequently. It turns out that getting a wave of feedback instead of a trickle helps lift workers’ job performance, but only if workers have enough time to assimilate the information before the next wave of feedback rolls in.

Casas-Arce conducts most of his research on how people respond to incentives and how to best design incentive programs. The topic is increasingly important to accounting professionals as companies increase the link between performance and compensation, and the complexity of compensation systems by including non-financial as well as financial performance measures in their analyses.

“We’re starting to realize that these non-financial performance measures have additional information that financial measures don’t,” Casas-Arce says. “For instance, if we can measure customer satisfaction, we may think a satisfied, happy customer is a customer who is going to come back. Those kinds of measures may help us predict future financial performance. They help us identify issues the company may face today that will drive performance in the future.”

Observing feedback and workers’ performance variations

Casas-Arce got the chance to experiment with feedback and worker performance when he and his team helped Multiasistencia, a European company that coordinates work between insurance companies’ customers who need repairs and the plumbers, masons, and other professionals who make the repairs. Multiasistencia wanted to design an incentive plan for their network of independent contractors that would improve customer satisfaction and reduce the percentage of the company’s detractors — customers who said they would not refer the company to others.

Key to making the incentive plan work, however, was giving the repair professionals feedback so they could learn from mistakes and improve their future performance. The next steps were determining whether better performance would come from giving monthly or weekly feedback, and whether it would come from giving aggregate information — the total number of a professional’s detractors during the period — or detailed information — the score for each service the professional completed during the period.

“We oftentimes think that once you design the incentives, people respond efficiently, but that’s not quite true in practice,” Casas-Arce says. “People require some help to respond effectively to incentives, be it through mentoring or providing them with the information they need to do their jobs. In this case, we felt that the professionals needed some feedback so they would be able to learn from it and improve performance. We were asking to what extent would people use information efficiently, and what was the best way to communicate that information?”

Views of economists, psychologists differ

Economists and psychologists have long offered different answers. Economists traditionally tend to assume that people act rationally and respond best when they have more information. Psychologists traditionally highlight people’s biases, such as a tendency to put more weight on recent or salient information than on their full range of information.

To conduct the research, Casas-Arce’s team collected detailed performance data on 800 professionals who did repairs for Multiasistencia as the company transitioned to its new incentive plan. Specifically, the data included customer satisfaction ratings for each job each professional performed.

The company tracked customer satisfaction ratings for four months, establishing a baseline for that metric. It then told the repair professionals about the new metric, the new monthly bonus plan, and their performance in the previous month.

Then the research team’s experiment began.

To determine which frequency and level of feedback would result in better performance, the researchers randomly assigned professionals to one of four groups. Over the next three months, one group received weekly, aggregate feedback; another weekly, detailed feedback; another monthly, aggregate feedback; and another monthly, detailed feedback.

For all groups, bonuses were based on their aggregate scores at the end of each month. In the economists’ view, professionals who were told of their scores each week would perform at least as well as those who were told of their scores monthly. In the psychologists’ view, professionals who were told their scores each week would focus more on the latest week’s score than on their previous scores.

The new feedback and incentive plans worked overall. After three months, all the professionals had improved their performance compared to before the feedback and incentive plans began.

Professionals thrive with monthly, detailed feedback

The specific results, though, came as a bit of a surprise to Casas-Arce and the company. Contrary to their initial expectations that people receiving frequent waves of information would behave rationally, they quickly saw that the professionals who performed best were the ones who received their feedback monthly, or less frequently, and received it in detail. This group’s share of detractors dropped the most, to 8.4 percent from 14.2 percent before the feedback and incentive plans started. The share of detractors for all three other groups dropped less dramatically, to between 10.5 and 11.2 percent from between 13 and 14.5 percent before the plans started.

Those results tell Casas-Arce that psychologists have a point: Infrequent feedback improves professionals’ ability to process the information, identify true trends, and make better decisions. When professionals receive frequent information, no matter whether it is aggregate or detailed, their performance deteriorates.

“People pay a lot of attention to the last piece of information they receive, and they neglect previous evidence they may have received,” he says. “When you give these professionals detailed information monthly, they have a lot of information they can process, and they’re able to improve performance. If you’re giving them very frequent information, they’re basing their decisions on little information, because they’re only paying attention to what happened last week.” Perhaps people lack the time or organizational skills, but for whatever reason, he says, those who get frequent information fail to put early and recent information together.

The researchers looked at other possible explanations for why those receiving weekly information could have performed worse than those receiving monthly information. One possibility was that if professionals learned early in the month that they had scored poorly, they would abandon hope of getting the bonus and slack off for the rest of the month. That would be rational, but the researchers found that professionals who learned they had performed poorly in the last week of a month also performed poorly in the first week of the next month, when new bonuses became available, indicating that the professionals were still thinking about the recent, negative feedback.

In addition, professionals who scored poorly in the last week of the month but received monthly feedback performed better in the first week of the next month, which told Casas-Arce that these professionals could put their feedback into context and didn’t overreact to the negative feedback. The researchers also found that regardless of the professionals’ ability levels, education levels, or length of time with the company, all responded similarly to their feedback.

The research persuaded Multiasistencia to give all its professionals monthly, detailed feedback. Casas-Arce’s team tracked results for three more months, and the professionals’ performance continued to be higher than before the company introduced the feedback and incentive plans. More significantly, professionals who had been receiving weekly feedback improved their performance under the monthly system, and their performance no longer deteriorated after they received negative feedback.

Casas-Arce says that although he believes the results are generalizable to many other industries, the word “frequent” may mean different things in different contexts. For him, it would be interesting to see how other workers, such as those whose tasks take longer or are repeated less frequently than the repair professionals’ tasks, respond to different feedback periods and detail. The research also could be extended to other contexts, such as financial markets in which investors often react to companies’ latest financial news instead of long-term performance. This has driven firms such as McDonald’s and other retailers to stop providing monthly sales reports to investors.

The bottom line

Finding that professionals perform better when they receive less frequent but more detailed feedback has these lessons for various stakeholders, Casas-Arce says:

• For workers receiving feedback, be aware that people have biases in the way we process information. Try to mitigate your own bias toward recent or salient information by stepping back, looking at all the information you receive, and then deciding.
• For accounting professionals, pay attention to how you present information. Think about how frequently you want to provide it, because recipients might overreact to a small piece of information that may be irrelevant in the long run.
• For companies and organizations, be cautious about how you handle today’s massive flows of information. Even though computers can crunch the data and you need to respond to opportunities and threats, people still have limited ability to process vast quantities of information. Taking time to review a broader swath of information is a good exercise that enables you to see the big picture.