Can you conduct Online Research among U.S. Hispanics?

October 16, 2011 Author: Mario X. Carrasco

We frequently get asked this question. The answer is a qualified “yes”. I say qualified because the online U.S. Hispanic population is decidedly diverse and having an understanding of its composition is required before engaging in any market research among this population. Before setting quotas for a particular survey you need to first ask yourself which Hispanic population are you trying to read. I’m often surprised by requests to test Spanish language creative among acculturated, non-Spanish speaking Hispanics or requests that assume a homogenous whole.

As you know, the U.S. Hispanics population in the West and South is very different (read higher numbers of less acculturated and Mexican descent) than on the eastern seaboard (higher acculturation levels and concentrations of Caribbean Hispanics). These differences are very visible when conducting online research since internet penetration can be as low as 45% among unacculturated Hispanics in the West to as high as 80% among acculturated Hispanics in the Southeast. I’ve helped build several large U.S. Hispanic online panels and have noticed some common characteristics among them:

U.S. Hispanic Panels skew eastward – Most panels mirror internet penetration and the further West you go the lower the Internet penetration levels are.

Unacculturated Men in the West are difficult to get online – Fortunately, much of the research we’re asked to conduct is focused on female primary grocery shoppers but, for any nationally representative U.S. Hispanic study, men in the West will likely be the last quota to close. Panels typically skew female so a male quota in the West has two trends working against it.

U.S. Hispanic Panels skew young and single – Hispanics are younger than the general population so perhaps this age skew is less pronounced on Hispanic panels than General market ones but it must be acknowledged. Also, approximately 49% of U.S. Hispanics 18+ are married while the population found on panels has a higher concentration of singles.

That said, it is possible to create a nationally representative study with carefully constructed front-end quotas that attempt to replicate the offline U.S. Hispanic population of 50.5 million drawing from the roughly 32 million of them that can be found online. While not ideal, this method for achieving representativeness is equal to or better than current off-line options since landline usage has plummeted and intercept work requires enormous base sizes spread evenly across the country to achieve representativeness.

Once geography, language usage, country of origin, education and income are controlled for the results from online research match other sources in most areas of interest. The results for any survey conducted online, however, will over-index the overall population for activities such as online shopping and social media usage but will be consistent for behaviors that are not tied to Internet usage such as food preference, personal care product usage, home care and other CPG categories.

Therefore, should you need a question answered about your U.S. Hispanics customers the online environment offers a multitude of research options. The assumption that a representative research study among Hispanics is not possible wholly online is starting to fade away as Hispanic savvy Fortune 500 companies begin to conduct all of their Hispanic research online, including among the unacculturated segment. The ubiquitous nature of the internet and sharp rise in smart phone ownership among all acculturation levels is making online research among Hispanics the mode of choice for companies looking to research this booming market.