Portland Marketing Analytics (PortMA) | Portland, Maine

How to Deal With Limited Demographics in Your Sample

Tactical vs. Strategic DataDealing with a limited demographic focus is to be expected when you conduct research. After all, no one is interested in everyone’s opinion on a product. Generally, you have a target consumer, whether that is an age range, gender, or household income. The issue arises when that demographic becomes too limited and you are left scrambling for a sufficiently large sample  from which to work.

So, what’s the best way to address this problem? As far as I can tell there are two basic solutions, neither of which are ideal.

Solution #1 – Make the target bigger

Simply take your least important parameter for your target and loosen it up a bit.

The difficulties with this are twofold.

  • First, you have to be able to convince everyone involved that this restriction is less significant, which can be no small matter.
  • Second, you may find, in your analysis, that the group you broaden was significantly different from what you originally targeted. By them, you either will have closed data collection or will have to extend it to ensure accurate results.

The key in these situations is how you leverage the data.

Figure out why this group that is close to, but not exactly the target demographic, is different, and why that is relevant. What does it indicate about the product or the target consumer that an age difference of 10 years, for example, would completely alter perception?

Solution #2 – Utilize multiple samples

The second, seemingly simpler solution, is to diversify your sample by using multiple sampling companies.

This, however, introduces the potential double sampling. It is possible that someone will participate in more than one event.

Not only that, how do you explain if your two sampling companies have wildly different results? Your analysis must now include why these groups are different and, as mentioned above, figure out why both are relevant.

This methodology also costs more for data collection.

How to prepare for limited data collection

There is no easy solution to this problem. I believe the best approach is to assess it on a project-by-project basis.

The question you have to ask yourself is,”How important is my targeting versus how concerned am I about the inherent costs of diversifying my sample base?

Photo Source: https://www.flickr.com/photos/mikecogh/

Share Your Thoughts...
Tell us what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar for free here!
Comments are moderated and may not appear immediately...