Snap’s diversity report reveals that the company has made very little progress in its diversity goals since last year, with 47 percent of its employees being white and 65 percent male. There was a tiny increase in women’s representation across the company (32.9 to 33.1 percent) and a similarly small increase in underrepresented racial groups among leadership (13.1 to 13.6 percent). Meanwhile, Hispanic / Latinx representation decreased from 6.9 to 6.8 percent, and Asian representation in leadership decreased from 16.5 to 14.3 percent.
The company is setting new representation goals, which it aims to meet by the end of 2025: increase underrepresented racial and ethnic groups across the company from 18.7 to 20 percent, increase women in tech roles from 16.5 to 25 percent, and increase women and racial and ethnic groups in leadership an additional 30 percent each, from 13.7 and 17.4 percent, respectively.
Last year’s diversity report was Snap’s first, and it was released in a turnaround from CEO Evan Spiegel’s announcement a month earlier that the company wouldn’t be sharing diversity numbers, saying he was worried “these disclosures normalize the current state of the tech industry.”
Snap also faced criticism last year for its Juneteenth filter, which included a “smile to break the chains” mechanism that was largely received as culturally insensitive and offensive. Oona King, Snap’s head of diversity and inclusion, apologized for the filter, saying that while Black Snap team members were involved in the creation of the filter, they had failed to consider the context of it being used by white people. Other Snapchat filters have been similarly criticized, including a Bob Marley filter and an anime-inspired filter that were both released in 2016.
One of the more concrete goals mentioned in Snap’s report is to develop a more inclusive Snapchat camera. Acknowledging the racist history of photography, the company is consulting with outside experts “to develop technology that counteracts bias.” This includes adjusting the front flash for the selfie camera and adding ways to correct brightness and exposure after a photo is taken, according to Axios. Snap says it is expanding the training of its algorithms with more diverse datasets to better recognize and capture images of non-white faces, as well as removing biased image adjustments like slimming people’s noses.