The Hispanic market is one of the most critical targets today and will continue to be even more important into the future. The approximately 50 million Hispanics in the U.S. today will become nearly 130 million by 2050. The purchasing power controlled by Hispanics, currently $1.2 trillion, is growing faster than that of any other population group. As their importance to marketers and advertisers has grown, the need for accurate insights – insights that cut through the stereotypes – has grown exponentially.
As their importance to marketers and advertisers has grown, the need for accurate insights – insights that cut through the stereotypes – has grown exponentially.
Many researchers, particularly those who are accustomed to conducting research among broad consumer targets, were historically reticent to conduct research among Hispanics. There was a time not so long ago when many market researchers avoided including any Hispanics in surveys except those who were highly acculturated. Commonly heard refrains included, “It’s too expensive,” “It’s too difficult,” “You can’t get a representative sample of less acculturated Hispanics,” “They don’t give accurate survey responses” and “The interviews are too hard to validate.” Equally troubling was the little recognized or acknowledged fact that Hispanics, along with other ethnic populations, were often dramatically underrepresented in surveys that purported to represent the “general market.” Marketers who requested of research vendors that their samples be ethnically balanced found themselves paying large surcharges for the privilege of obtaining representative samples.
Times have changed. Today, many researchers are including less-than-fully-acculturated Hispanics into their samples and even conducting many surveys specific to Hispanic targets. Thanks to the ubiquity of the Internet, online research panels and Hispanics’ strong usage of mobile devices, many of the old excuses for avoiding Hispanic research are increasingly an artifact of the past. Numerous researchers and online panel providers have scrambled to fill the once sparsely populated space with some supplementing general market capabilities with Hispanic proficiencies and others emerging as specialists in this sector.
In order to accurately represent the voice of the Hispanic consumer, the industry at large needs to address topics that range from what, precisely, we mean by “Hispanics” to how we can ensure that our research approaches are nuanced to fully capture the truths that we seek.
What actually constitutes the Hispanic market? The broad definition of Hispanics includes, at one end of the spectrum, individuals who may (or may not) have a Hispanic surname and were born in the U.S., perhaps even being second-generation Americans. Individuals within this group may not speak Spanish, may not identify with Hispanic culture and are unlikely to have any interest in Hispanic media. Are these individuals Hispanics, despite the fact that their behaviors and attitudes largely mirror those of non-Hispanic, general market individuals? We might say no, but those who quote statistics about the size and vitality of the Hispanic market include them in their definitions.
On the other end of the spectrum are individuals who are recent immigrants, who spend their lives surrounded by other Hispanics, are less participatory in American life and who speak little or no English. These individuals have historically been difficult to reach through standard survey methods. Based on their lifestyles and consumption habits, they are of less interest to certain marketers, but many would be missing an important piece of the puzzle were they to ignore this relatively large group. Unlike the more assimilated Hispanics who often act like general market consumers, there is simply no way to understand this group without talking to them. And surveys that do not include or underrepresent them cannot be accurately presented as reflecting the U.S. Hispanic population.
Over half of unacculturated Hispanics were born outside the U.S. while nearly half of biculturals have been in the country more than 20 years. Both bicultural and unacculturated Hispanics are more likely to identify themselves as equally American and Hispanic.
When classifying Hispanics for conducting surveys and when developing advertising and marketing strategies, marketers and researchers think in terms of acculturation levels. But what are the specific metrics upon which acculturation is based? And where are the lines between unacculturated and bicultural or between bicultural and acculturated? And should fully acculturated Hispanics be categorized as Hispanic at all?
An informal survey of advertisers who target both Hispanics and general market consumers clearly indicates a lack of consensus on these questions. One large CPG marketer uses definitions that are styled from the U.S. Census with questioning based solely on the language spoken at home. Using these definitions, 28 percent of U.S. Hispanics are defined as Spanish dominant, 20 percent as English dominant and 52 percent as bilingual. More typically, many advertisers use “point” systems, with points accumulated based on dimensions that may variously include country of origin, length of time in the U.S., language spoken at home and/or out of home, language used for reading books and magazines or watching TV and the culture with which they most strongly identify.
How these “points” are accrued, and what constitutes an unacculturated Hispanic versus a bicultural versus an acculturated Hispanic vary quite significantly across different marketing organizations. While most advertisers divide the Hispanic population into three groups, a few use four: low, medium, high and fully assimilated. Some don’t count the acculturated or highly assimilated as Hispanics at all, but others do. Some focus only on the middle group, the biculturals, believing that the acculturated group is indistinguishable from the rest of the general market and that the unacculturated are too far removed from the mainstream to drive significant results for their brands.
It’s no wonder that we can’t agree on how to think about marketing to Hispanics when we can’t even agree on common definitions
It’s no wonder that we can’t agree on how to think about marketing to Hispanics when we can’t even agree on common definitions. It’s probably unrealistic to think that we can form a consensus as to how we define Hispanics for research or marketing purposes, but it is a conversation that is worth having.
When it comes to conducting research among Hispanics, the online panel providers have really upped their game in the past few years, responding to marketplace demand. Less acculturated Hispanics are still underrepresented in most online panels, and even Hispanics overall are not present in the numbers that their overall population incidence warrants. But thanks to the diligence of panel providers, they do exist in research panels and in numbers that enable the inclusion of these important voices to those who would make a concerted effort.
However, the fact that Hispanics, particularly the less acculturated, are still dramatically underrepresented in panels and other research communities means that it still takes a special effort to include them to the degree that their population incidence requires. Consequently, it remains critical that quotas are established to ensure that unacculturated Hispanics, however that is defined, are adequately represented in surveys. If these quotas are rigorously adhered to, my experience suggests that online panel composition has evolved to the point that we may no longer need to supplement online panel surveys with costly in-person/mall intercept methodologies.
Once we have contacted the Hispanic consumer and requested their survey participation, ideally with the invitation presented in both English and Spanish, it is important that the survey experience be sensitive to the cultural differences that exist. It has become common knowledge that Hispanics are relatively heavy users of mobile technology, so providing mobile-friendly surveys is critical. From keeping in touch with friends and extended family to staying current with new music and artists to searching for retail locations and bargains, Hispanics make good use of their smartphones. In fact, because a single laptop computer is often shared by an entire family, an individual’s smartphone is often the one electronic device on which he or she fully relies at a personal level.
Not surprisingly, the mobile device is the go-to platform for Hispanics when it comes to survey participation. One panel provider reports that 65 percent of Hispanics invited to participate in an average study attempt to access it from a mobile device. Whether or not the survey is mobile-friendly – and how mobile-friendly it is – inevitably will have a strong effect on their participation or lack thereof.
Not only must surveys be offered on mobile devices, the Hispanic consumer must be provided with an easy way to complete the survey in either English or Spanish. When given a choice of languages, it’s not surprising that nearly all unacculturated Hispanics choose Spanish. While bicultural Hispanics are generally quite proficient in English, even among this group, it is common for half or more to choose Spanish. Surveys that aren’t offered in Spanish, or those that make it difficult to switch from an English language to a Spanish language version, will potentially be off-putting to large numbers of Hispanics who prefer to respond in their native tongue.
A small number of acculturated Hispanics opt for Spanish language surveys while bicultural Hispanics are nearly equal in preference between Spanish and English.
And speaking of the interview, Hispanics tend to report more favorable survey experiences than do non-Hispanics. Using one common scale of survey satisfaction, 67 percent of Hispanics report high satisfaction and only 10 percent low satisfaction, while just 52 percent of non-Hispanics rate the experience favorably and 15 percent consider it unfavorably. This despite the fact that it takes Hispanics on average slightly longer to complete the same survey as it does non-Hispanics. While researchers should be cautious about taking advantage of the good-natured and cooperative dispositions of Hispanic respondents, this group is, as a whole, more likely to tolerate long surveys and less likely to terminate mid-survey.
Hispanic respondents express stronger survey satisfaction than do non-Hispanic respondents and are less likely to be dissatisfied with the survey experience.
This “agreeability” phenomenon is also commonly observed as an inclination for Hispanics to provide more favorable responses across a wide variety of survey topics. In research across a range of consumer product categories, Hispanic consumers tend to have the strongest favorability biases on those questions that are the most subjective. For example, the cultural biases are strongest on emotionally-driven attributes (e.g., “A brand I feel good about serving to my family”) and relatively weaker on attributes that have a more tangible basis (e.g., “A brand that is all natural”). When it comes to behavioral data, Hispanics are far more likely to report inflated future purchase intentions for brands than they are to over-report recent past purchasing. As with other Hispanic dynamics, this phenomenon is most pronounced with unacculturated and less evident as acculturation levels rise.
Hispanics’ ratings of brands are significantly more positive than non-Hispanics, particularly on emotional attributes and when reporting future purchase intentions. Unacculturated Hispanics provide the most positive ratings, but even acculturated Hispanics skew positive compared to non-Hispanics.
It’s not enough to identify patterns in favorable response bias among Hispanics. As researchers, it behooves us to find ways to address them in order to obtain the most accurate possible survey data. Some techniques that can be effective in reducing favorable response bias include:
- Using scales that require the respondent to choose between brands. Rather than ask for ratings of each brand on a 10-point scale, ask the respondent to select the one brand that is best on a given attribute, the one that’s second best, etc.
- Using analytic frames of reference that adjust for overall response. For example, reporting metrics that reflect the extent to which a particular brand’s score is above or below the scores provided by that respondent for competitive brands. If Brand A scores a 72 percent and the five-brand average is 65 percent, the brand’s relative score would be reported as a +7.
- Asking respondents to “prove” that they have, for example, seen a particular ad before. You could ask, “Have you seen this commercial while watching TV?” But it’s better to instead show a brief portion of the ad and ask respondents to describe what else they remember from the ad.
- Using pre/post measures among the same set of respondents. Instead of asking consumers whether an ad would make them buy the brand, construct a scenario in which they choose from a set of the brands. Then, after a commercial exposure (whether in the real world or in the laboratory), see if their preferred brand has changed.
Additionally, Spanish-language surveys must be authentic. Often, translations come across poorly. The sense is that Google translations or other non-human tools were used rather than real Spanish speakers. Overly formal language that does not reflect how people actually speak is both confusing and a turn off.
When constructing surveys for Hispanics, the old “KISS” (Keep It Simple, Stupid!) adage should be used for guidance. Ideally, questions should be short and to the point as should response categories. The more that’s said, the greater the need to provide translations and the greater the opportunity to have respondents misconstrue the questions or the instructions. This advice is equally applicable to surveys that are designed wholly for non-Hispanic targets. In this case, when an optimal survey is designed for Hispanics, chances are that you may be at the same time improving the instrument that you use for your general market surveys.
We’ve certainly made advancements in the past several years in terms of our ability to include the voices of all types of Hispanics in quantitative surveys. Further, we have better, more proven protocols to ensure that we gain meaningful insights from the questions that we ask. It’s still not as easy to gain a representative sample of Hispanics, covering the full spectrum of acculturation levels, as it is to use convenience samples that seriously underrepresent them. And we need to continue to be cognizant of the cultural differences that affect their survey-taking experience and responses. However, for researchers who recognize the importance of hearing the distinctive voices of all Americans, those old objections — “Too difficult,” “Too expensive,” etc. — have become less valid. Accurate insights that reflect Hispanic opinions, attitudes and behaviors are well within our reach.