Corporate CEOs who apologize to customers by video had better feel sorry -- or they’ll end up sorry, study finds

August 15, 2013

For more information, contact: Ben Haimowitz, 212-233-6170,

Aug. 15, 2013 - When Apple CEO Tim Cook apologized in April to the Chinese people for the company's iPhone warranty and repair policies, he did so via a letter posted on the company's Web site, the same method used last year in another mea culpa to Apple customers about the company's map apps. In an era when CEO apologies have developed into an emerging art form on YouTube, shouldn't Apple be doing more than posting letters of regret?

Probably not, new research suggests. CEO apology videos, it finds, are hazardous undertakings that can easily prove counterproductive -- and for reasons that were hinted .at almost 150 years ago by none other than Charles Darwin when he hypothesized that some facial muscle actions associated with strong emotion are beyond voluntary control.

A paper that was presented at the annual meeting of the Academy of Management (Orlando, Aug. 9-13), employs advanced technology to bring Darwin up to date in analyzing 29 corporate apologies over a  recent five-year period. It finds that, "while verbal content of an apology has relatively minor effect, inappropriate facial emotions can be greatly detrimental to stock market performance....A brief leakage of disgust, or a misplaced smile, can damage fragile relationships and send stock prices plummeting."

Leanne ten Brinke of the University of California, Berkeley, who carried out the study with Gabrielle Adams of the London Business School, compares decisions to put CEOs before the camera to the decisions defense lawyers make on whether to put defendants on the witness stand. "If you're going to go before a camera and say you're sorry, you'd better feel sorry, or your facial muscles may betray you and only make matters worse. Can a top corporate executive who doesn't feel particularly regretful be coached into looking that way convincingly? Probably not. How do you achieve in a day or two what may take professional actors years to master?"

The research sample consists of all corporate transgressions from 2007 through 2011 for which apology videos were publicly available along with company stock-performance data sufficient to gauge to what extent returns departed from normal patterns in the wake of apologies. Some 29 events met the data requirements including such notorious ones as the 2010 BP oil spill, JetBlue's 2007 operations breakdown that caused severe travel disruptions for its passengers, and Mattel's 2007 recall of more than 18 million products because of possible dangers they posed to children. In 16 cases the video apologies were delivered by a CEO and in 13 instances by another company executive.

Unsurprisingly, the research revealed that, on average, companies sustained negative returns of 12.78% over the course of the 30 days before the apology, a period that in most cases included the foul-up that made the mea culpas necessary. The downward trend continued the day of the apology and the following day but improved to 3.78% below the long-term norm in the 30 days following the video.

Of greatest interest to ten Brinke and Adams, though, was how the apologies affected stock performance on apology day and apology day plus one, when the market's reaction to the video was likely to be most clear -and, specifically, whether differences in the audio and visual components of the videos might be related to differences in financial returns on those days. Why, for example, was the apology of Sony CEO Kazuo Hirai to Playstation subscribers accompanied by a 2.37% return above the norm on video-release day, while that of Netflix CEO Reed Hastings to his customers brought negative abnormal returns of 7.44% on release day and 9.70% the day after?

The authors found that the only verbal content that was related to financial outcomes, and just slightly, "was the inclusion of a remorseful statement, although this was only marginally significant. This suggests that other verbal content of an apology such as promises to repair or offers of forbearance… do not have a strong effect," they write.

In striking contrast, analysis of the facial expressions of the apologizing executives with the sound off to avoid potential bias introduced by speech content or tone of voice uncovered highly significant effects. To quantify phenomena as elusive as facial expressions, the researchers employed a method that has been developed in other research on deceptive behavior, whereby video content is broken up into 30 still frames per second and trained coders scan each frame, the lower and upper face separately, for expressions conveying happiness, sadness, and disgust. "Facial hemispheres are coded independently," the authors explain, "because coding full-face expressions is not sufficiently sensitive to capture the complexity of emotional facial movement…[E]motional signals of deception are rarely full-faced and instead are seen in particular muscles that are less under volitional control."

The researchers quantified inappropriate affect by having coders count frames to determine the duration of disgust and happiness expressed by the upper or lower face in contrast to expressions of sadness. As ten Brinke and Adams explain, "The apologizer should neither show nor feel happiness about the transgression committed, or the plight of any victims. Although signs of disgust might indicate that the apologizer feels distaste for the organization's transgressions, previous research had found that expressions of disgust do not occur in genuine expressions of remorse and are instead a better signal of insincere remorse."

Evidence of happiness, as expressed by smiles, and of disgust, as evinced by mouths turned down, were strongly correlated with negative stock returns on the day of the apology and the following day, with less than a 5% possibility that the relationships were due to chance. In contrast, evidence of sadness "protected the firm from further damage to stock price."

While these outcomes were true for the sample as a whole, the most dramatic results were for CEOs, a result that the authors ascribe to "third-party perceptions that CEOs personify an organization's values and that inappropriate affect (i.e., perceived lack of remorse) on their part makes members of the public less confident that the organization will avoid similar future transgressions."

"Obviously, not all stockholders watch an apology," ten Brinke adds. "But those that do see it and are struck by lack of remorse evidently suffice to have a substantial negative effect on market prices, particularly when the apology is by a CEO. As reports of their reactions spread through the market and create concern among other investors, the stock price will take a hit."

The paper, entitled "Inappropriate Facial Expressions in Public Apologies Predict Investor Confidence," was among several thousand research reports at the Academy of Management annual meeting, held in Orlando from August 9th through 13th. Founded in 1936, the Academy of Management is the largest organization in the world devoted to management research and teaching. It has some 19,000 members in 110 countries, including about 11,000 in the United States. This year's annual meeting drew more than 9,000 scholars and practitioners for sessions on a host of subjects relating to business strategy, organizational behavior, corporate governance, careers, human resources, technology development, and other management-related topics.