Corporate ads said Black Lives Matter. But did brands deliver on the promises made during the summer protests?
Black Lives Matter galvanized many companies to take a stance against systemic racism and white supremacy. Diversity and inclusion became a CMO priority, and a commitment was made to re-evaluate the ads brands spent billions of dollars a year promoting. The focus was on creating "un-stereotyped, meaningful, and effective communication," Tamara Rogers, CMO of GSK said, "to drive the cultural change we need".
To learn more about how marketers responded to Black Lives Matter and the stereotypes that prevent marketers from producing inclusive ads, we used our creative excellence technology to analyze 2,738 image and video ads deployed in the US in 2020 by some of the world’s best known CPG, beauty, and alcohol brands.
Using computer vision, we measured -- on a month by month basis -- the frequency of visual cues that appeared in each creative, including people, diversity, setting, etc. Based on our analysis, there are three things marketers should be aware of when attempting to increase the diversity and inclusivity of their ads:
Below are 17 insights that marketers can use to evaluate their own creative content.
There is a clear pattern in the data: when white people feature more in creatives, Black people feature less (and vice versa).
Over the course of 2020, the number of ads featuring Black people fluctuated, but experienced a progressive increase from 42.8% in January to 52.7% in December.
As the number of ads featuring Black people increased, the number of ads featuring white people fell from 89.9% in January to 74.2% in December.
Black people featured in approximately 46.3% of ads during the first four months of 2020. This number dropped to 46% from May to August, only to rise to 53.6% during the months of September to December.
Ads featuring Black people peaked in September at 62%, a small increase over the number of white characters (60.1%). Further analysis (see below), suggests that this could be due to the rise in sporting events (i.e. NBA Playoffs, start of the NFL season).
Digging deeper into representation, it’s clear that gender plays a role.
Over the course of 2020, Black women featured approximately 65% more than Black men. There are no months where Black men appeared in more ads than Black women.
Presence of Black women increased 68% from May to June, while representation of Black men stayed roughly the same from May to June (around 20% of all ads featured Black men), suggesting BLM had no impact on marketers' creative decisions when it came to casting Black men.
Further analysis of representation found that speaking roles followed a similar pattern, regardless of ethnicity. With the exception of April, ads featuring characters in speaking roles did not diverge throughout 2020.
Over the course of 2020, Black people featured in speaking roles more than white people. There are, however, two points of note. First, speaking roles for Black people fell during the months of protest (from 32.14% in April to 7.25% in May and 14% in June and July). Second, Black people had more speaking roles than white people from August to December (after the BLM protests).
With the exception of April (Covid positively impacted speaking roles for Black people - for an explanation as to why, see insight #15), ads were more likely to feature Black people in speaking roles when the Black person was in an ad classified as racially stereotyped (see below for classification details).
A gendered analysis of speaking roles found that male characters spoke more frequently than female characters.
With the exception of April, August, September, and December, white male characters were more likely to be cast in speaking roles.
Over the course of 2020, Black men were about 6% more likely than Black women to feature in speaking roles (despite 65% less screen time).
Increased representation of women and Black people was expected to come at a price: more stereotypes. Three reasons were behind this hypothesis:
To evaluate for racial stereotypes we categorized the following visual cues as characteristics typically associated with Black people -- due to racist bias and physical stereotyping -- shirtless men, people doing physical activities (and by extension people watching sports wearing team kits). The use and frequency of racial stereotypes in the ads we analyzed followed two patterns.
June, September, and December were months in which representation of Black people increased. These were all months with corresponding increases in racial stereotypes. For example, representation increased 15% in June and stereotypes more than trebled.
White people were more likely to feature as skiers and golfers while Black people were more likely to feature as basketball players. Black people in stereotyped roles nearly trebled in August, which is when the NBA Playoffs start. Interestingly, this is also the month that ads featuring a single Black person surrounded by white people (i.e. the token Black friend) increased nearly 20x.
Over the course of 2020, white characters dominated the leadership roles. April was the only month with ads featuring more Black people in leadership roles than white characters (10.71% vs 6.73%).
Ads featuring Black people in leadership roles fluctuated between 0% and 2.5% of ads. Given that approximately 50% of ads in 2020 featured a person of colour, 1% of all ads in 2020 featured a Black person in a leadership role.
It appears that brands did not deliver on the promises to produce diverse and inclusive ads.
Digging deeper into why Black people featured in leadership roles more frequently than white people on two occasions reveals that Covid played a role during the most significant increase.
The onset of Covid-19 saw brands focus on and champion Black doctors. Unfortunately, this positive uptick was not maintained.
As female representation increases so too does the frequency of sexualized female characters. Female characters are more likely to be cast in a sexual portrayal when representation increases (e.g. May, August, and November).
Over the course of 2020, 7.75% of ads featuring Black women portrayed them in a sexualized manner, compared to 6.55% of ads featuring white women. Ads featuring sexualized Black women increased nearly 7x in June, which is concerning when compared to the complete lack of Black people cast in leadership roles during June, July, and August.
Our analysis revealed some surprising creative decisions and trends in 2020. The Black Lives Matter movement catalyzed promises of change but challenging unconscious bias and removing stereotypes from the way we represent and portray different people are clearly difficult tasks.
The data revealed that marketers did not deliver on the promises they made during the months of protest. Rather, Black people were less represented and speaking roles and leadership roles both fell. It appears that we have a bias towards including Black people when sport is a theme or representation is a box-ticking exercise, leading to appearances of the "token Black friend."
Opportunities exist. Brands can and should make concerted efforts to increase the diversity and range of skin tones that feature in their ads. Women continue to be cast as background or secondary characters. They need more speaking time. To use advertising to create cultural change and challenge social expectations, Black people must feature more in leadership roles.
This is part one of a two-part analysis into the state of diversity and inclusion in advertising. On March 8th, we will be releasing a detailed analysis of how women are portrayed in advertising. Sign up here to get notified when it's released.
To see if our technology can help you on your mission to unstereotype your ads, get in touch.