I hate social media research because: there are no demographics #1 #MRX

Research Team

I recently wrote a blog post citing ten of the biggest complaints about social media research. Today I address complaint #1.

Social media research does not include demographic data.

I’ll admit right off the bat that this is a commonly raised concern. Coming from traditional research where every single survey complete and focus group attendee is associated with at least a couple of demographic datapoints (age, gender) and possibly even 8 or 10 variables (household size, number and age of children,income, education,  region, religion, race, hispanic, language), it’s hard for experienced researchers to come to grips with datasets where they don’t know the stated demographics of even a few of the participants.

Unfortunately, in the case of social media research, the problem is more than a few records lacking a few demos. Indeed, a dataset may only have demos for 1% to 20% of datapoints. And worse, there might only be one or two demographic pieces of data, like age or gender, for those records.

But let’s not look at the glass half empty. This is definitely a glass half full moment. I am still astounded when I see a social media dataset where 20% of the data has demographic information. Really, when was the last time you wrote a tweet like this?

So think about it this way. If you NEED detailed demographic data, then your main methodology should be a survey of sorts. Social media research should be there to fill in the gaps, to add flavour, to show details that couldn’t be evaluated with the survey data. You must always choose the right method for your research objective and if that means doing a survey or a focus group so that you have demographic data, then so be it.

True or False? Mostly true.

5 responses

  1. While I agree that SM lacks demographic information, I disagree that we always make most use of demographic information, barring in segmentation studies. Tables of 300 pages is very common in research, regardless of type of studies however I always wondered that for non-segmentation research projects, even if you were to find meaningful differences across demographics, would your marketing team refine its marketing material/positioning? I have seen several projects like ATU, message recall etc. where vendors provide 300 page output in word document showing differences across demographics but why? If it is segmentation study, sure it is worth it but otherwise is it nice to have info or must have info? Are you really going to take action on such demo differences?

    If one is still hungry of demographics, it may not hurt to adopt google approach on sampling (extrapolating location based demographics).

    I am Manmit Shrimali, male living in toronto, non-hispanic, with prefer not to answer income and age 🙂

    1. LOL. I respect your decision to use the “prefer not to answer” option.
      Married, white, female

    1. And why didn`t you leave your full demographic information in your comment? 🙂

  2. […] read a blog post on the LoveStats blog today that referred to one of the most widely regarded critiques of social […]

%d bloggers like this: