Is market research still necessary in the age of social media? Part 2

By Gonzalo López Martí / LMMIAMI.COM

• Many a garden variety strategic planner will bump into a lame remark from some nobody or other on the twitterwebs and quote it like gospel.
• This is happening now.
• A lot.
• Social media is increasingly replacing market research in our line of business, whether we like it or not.
• However, as any statistician fresh out of college can tell you, a sample survey is not statistically accurate if it’s not harvested randomly.
• That’s why social media is pretty fuzzy from a scientific POV.
• Even fuzzier than those darned focus groups.
• SocMed only represents the extremes.
• The vocal, the loud, the idle, the trolls, the folks who have skin in the game or too much time on their hands.
• The very busy or very lazy happy middle is absentmindedly witnessing the whole thing from the sidelines and rarely bothers participating.
• To some extent, this statistical postulate even questions democracy as we know it.
• Jorge Luis Borges used to mordantly say, or so the myth goes, that “democracy is an abuse of statistics”.
• In the US, for instance, where voter turnout is alarmingly low, it could be argued that the appointment of elective posts is monopolized, even hijacked, by a certain minority with the spare time -or the willingness- to cast their vote.
• OK, back to marketing.
• But before a little more politics: another sound bite, in this case attributed to late 19th Century German statesman and warmonger Otto von Bismarck, goes something like “if you like sausages and laws, you’d better not find out how they make them.”
• The same applies to research, folks.
• And when I say “research”, I mean “quantitative research”.
• Qualitative research, aka focus groups, are a different story.
• Anyone who’s ever seen how a poll is conducted knows that it must be executed with absolute regard for randomness.
• For instance: the interviewer must place him or herself at the entrance of a broad-spectrum shopping mall and religiously interview every seventh patron that crosses his or her way.
• If possible, blindfolded.
• The interviewer cannot pick interviewees, which would represent a conscious or unconscious bias that’d make the poll invalid ipso facto.
• Every seventh patron must be approached.
• If the patron refuses, count to seven again.
• Still, the margin of error is high because no shopping mall in America has the proportion of ethnic and socioeconomic customer profiles to fully and truly represent the country.
• Hence, several shopping malls should be surveyed (or train terminals, public parks and so on and so forth) in different parts of the country, as many as possible, which rarely, if ever, happens.
• The logic is simple: if someone’s participation in a poll is not totally random or, worse, if he or she participates by his or her own initiative, his or her opinion taints the outcome, therefore rendering said poll inaccurate and invalid.
• Skin in the game kills the mandatory randomness of it all.
• Hence, ladies & gents, social media does NOT represent a valid reflection of public opinion or sentiment on a certain issue.
• Quoting tweets to back your personal bias does not account for market research.
• Hypothesis debunked.
• Perfect quantitative market research might not exist and it always carries a considerable margin of error but it still is waaaaay more accurate than so called “social listening”
• I rest my case.
• I mean, no hard feelings against “social listening”.
• It is ok to use the Twitbook as food for thought, the same way it doesn’t hurt every now and then to watch a dozen people talk about your product or campaign from behind a two-way mirror.
• It certainly lubricates the mental process.
• Yet neither of these settings have any statistical value whatsoever.
• By the way, you have just read the first and probably the last time a creative speaks out in favor of market research.