Commentary

How Defining Hispanic, Multicultural Ethnicity Online Will Impact Digital Marketing In 2018

In 2017, consumer demand for diversity and inclusion ignited a watershed in the evolution of multicultural marketing and research. Global brands publicly touted their commitment to these ideals, as seen by change agents like Nike, with the launch of the first mainstream sports hijab and Disney Pixar, whose animated movie, Coco, based on the Mexican holiday of Día de los Muertos, shattered the box office this holiday season.

But the ANA Alliance for Inclusive & Multicultural Marketing (AIMM) is not one to rest on these laurels. At the 2017 Multicultural Marketing & Diversity Conference of the ANA (Association for National Advertisers) in November, it announced a set of key priorities for 2018 designed “to help marketers increase their focus on multicultural marketing to help drive overall corporate growth.” The priorities range from creating a case for change built on qualitative and quantitative information to developing best practices on “Total Market.” 

advertisement

advertisement

All the priorities outlined by the AIMM are important and should be pursued by any marketers looking to drive overall corporate growth through multicultural marketing. However, as a digital insights firm, the priority related to data and metrics was of extreme importance to us as it states: 

The AIMM will lead an industry-wide initiative to identify, prioritize, and close existing data representation and quality gaps in multicultural segments. The effort will examine what impact cultural relevance has on digital campaigns as opposed to campaigns without cultural insights, regardless of language.

While a step in the right direction, this priority, purposefully written at a high level as it’s sure to evolve, would benefit from additional insight into how we’re defining the ethnicity of online users before we try to fix the lack of representation and quality gaps in multicultural segments. Different online platforms have different ways of defining ethnicity based on a variety of online behaviors. They are generally referred to as affinities or affinity models. 

Google and Facebook use complex affinity models.

With access to a tremendous amount of user data, Google and Facebook use sophisticated affinity models to define ethnicity online based on online behaviors. Their affinity models are based on factors like type of music streamed, type of content liked, language of web searches, etc. While the data points used to create the algorithm that feeds their affinity models differ, the approach is similar. 

DMPs and Publishers use fewer factors in affinity models.

Many DMPs (data management platforms) and publishers, however, have a different approach to affinity models. Focusing specifically on the U.S. Hispanic online user, some DMPs and publishers take a less sophisticated approach to ethnicity models and rely on language usage. There are a couple of issues with this approach as Spanish-language captures a small portion of Latinos in the U.S. and many Spanish speaking Hispanics may not be browsing the internet in Spanish. 

Defining ethnicity varies across affinity models. 

While these two approaches are by no means exhaustive, they do represent the two primary buckets ethnicity definition online tends to fall into. Defining how we are going to measure this factor is essential to ultimately addressing quality gaps in multicultural segments. While it may be nearly impossible to implement a blanket approach industry-wide, we can do a better job of adding more rigor to affinity models. Going beyond online language usage is a good first step. 

Furthermore, awareness of the fact that ethnicity definitions online are by no means set in stone is would benefit marketers. Asking questions and getting clarity on how your advertising partner is defining their audience can lead to better solutions. Sharing internal data with publishers can also lead to more robust models that can help the industry better define ethnicity online.

Next story loading loading..