That Starbucks is trying to lead a national conversation on race is fraught for so many reasons—among them, the fact that most of its leadership team is white. A new composite image of Starbucks's executives sums it up pretty well:
The image, made by Matthew Skomarovsky, a software developer and computer artist based in New York City, combines the headshots on the leadership page of Starbucks's website. Skomarovsky made the picture with a custom script that stacked the photographs, lined up the eyes of each face and took the median of each pixel from all the images.
Skomarovsky tweeted his image today, a few days after Starbucks CEO Howard Schultz announced a new campaign called Race Together that is aimed at opening up a dialogue about race between its cafés' baristas and customers. Schultz has encouraged baristas to write "Race Together" on the cups they serve customers to start conversations about race.
This Friday, the company will also release supplementary reading material in USA Today that includes conversation starters such as "In the past year, I have been to the home of someone of a different race _____ times."
From just a cursory glance at Starbucks's leadership page, Skomarovsky was confident about what he would find—15 of the company's 19 executives are white, and 13 of them are specifically white men. "I wasn't surprised at the result at all," said Skomarovsky, who co-founded LittleSis, a watchdog database that tracks connections between powerful people and organizations.
"I've done this before with Bank of America and other major corporations, and the result is usually pretty similar," he told me. "The face tends to look like a soft-looking, white, middle-aged male."
Skomarovsky said he made the image partly out of curiosity, but partly to make a point using the technology he already had
To stitch together the faces of the Starbuck executives, Skomarovsky used a script he wrote using the programming language Ruby. The script stacked together all of the headshots, lined up the centers of each person's eyes and then took the median of each pixel. To automate the process of lining up each face's eyes, Skomarovsky used OpenCV, an open-source library of programming functions that includes facial detection programs. He created the final image using ImageMagick, an open-source image processing program.
The process took about 15 minutes. "This is something you can do in Photoshop too, with just a little more manual effort," said Skomarovsky.
Skomarovsky said he made the image partly out of curiosity, but partly to make a point using the technology he already had. "I knew people were talking about Starbucks's relationship to race," he said. "I felt like this might be a fun provocation that ultimately points out that Starbucks is a very white-male dominated company, just like most other large companies in America."
It's a point well made—Skomarovsky's image is perhaps best summed up by Starbucks's newest signature drink, an Australian espresso drink called the Flat White.