Analytics: Turning the Multiplication Tables
By Lynn Russo Whylly at Media Post
Diversity marketers have the tools at their fingertips.
Pop quiz time:
What percentage of your customer base is multicultural? And what percentage of those consumers engage with you through your Web site and online advertising? If you shrugged and muttered “I have no freakin’ clue” then consider yourself among the majority. Although companies spend bucket loads of time and money on analytics to understand and predict the behavior of their general customer base (both on and offline), very few put the same amount of money and effort into understanding their multicultural customers.
Big mistake, pal. Here’s why: More than 79 percent of Hispanics, 75 percent of Asians and 56 percent of African Americans are online, according to the Pew Internet & American Life Project. With numbers like these, it’s almost certain you’re doing business with many of them. But by not understanding their specific needs and behaviors, you’re leaving money on the table.
Multicultural analytics is out there sporadically, but it’s not considered a best practice yet, mainly because it’s hard to substantiate against the bottom line,” explains Andrea Hoffman, founder and CEO of consultancy Diversity Affluence. While some companies have stuck a toe in the water, Hoffman says there’s too much awareness and not enough action. “They’re discussing this at multicultural conferences ad nauseam,” she says. “It’s time to move the needle. What we need are more case studies and less research and theoretics.”
To get to work, companies can begin with an opportunity assessment, profiling existing customers by ethnicity, says David Perez, CEO of multicultural market intelligence firm Latin Force. They should then augment this information with third-party data so they have a complete understanding of who their multicultural customers are and how they compare with their best customers and the general population. “In order to find more prospects like them, you first have to know what you have to start with,” Perez says.
Companies should ensure their data-collection efforts include diversity fields like country of origin and language preference. Name identification is one of the most popular tools for segmenting multicultural databases, says Devyani Sadh, Chief Data Officer at analytics consultancy Data Square. There is software available that can identify variations from an ethnic and nomenclature perspective through unique types of names and the way those names are structured, Sadh says. Lists of retail product returns plus law enforcement, health care and airline records can help with name identification. You can match your file against these databases where people have either deliberately or accidentally used different variations of their name, Sadh explains. You can see the combinations of the same name and then merge/purge them into a single name. After identifying the name, she suggests companies classify them according to ethnic background and add the person's gender.
A Web site registration page can also capture ethnicity, language preference and other cultural cues, Perez says.
Sadh cautions, however, to proceed with sensitivity. If you come right out and ask for their race, you may not necessarily get an accurate answer, she says. A more culturally sensitive and highly predictive alternative, she suggests, is to ask racially-biased questions, such as favorite foods and music groups, or holiday reminders they want to receive. In addition, she says, collecting U.S. Census or geographic data at the block level rather than the zip code level will ensure higher ethnic accuracy.
Once you have the right data in place, it’s time to create the intelligence. Sadh and Hoffman suggest starting off with a small test or pilot program, which is optimal, of course, for online. With the proper tools in your box, the next time you take the quiz, get ready to grab that cash back off the table.