Identifying Audiences in the Data Puzzle


If we handed you a puzzle piece, could you answer the following questions:

  1. How big is the puzzle – is it a 10 piece, 100 piece, 1,000 piece puzzle?
  2. What does the puzzle look like – is that blue the sky or the sea?
  3. Does this piece accurately represent the whole puzzle – is the whole puzzle a picture of the sea?
  4. How does this piece of the puzzle compare to other pieces – are the other pieces the same size, color or shape?
  5. Where does this piece fit into the puzzle – is it the center of the puzzle or upper left corner?

Identifying audiences within a data set is just like trying to figure out how a single piece fits into the entire puzzle. As marketers we now live in a world where using big data is the rule not the exception.

Big data is often thrown at problems without much thought as to how the data will be used – things like methodology of collection, aggregation and analysis are not considered in favor of accessibility. But in order to successfully develop audiences, you need to understand the data and where it comes from. Otherwise, how it’s used will become problematic and lead to misinterpretations that will eventually hurt your business.

In the era of big data, it is becoming increasingly rare that we receive a complete data set. We are not yet at a point where we are able to consolidate every internet cookie or every set-top box feed into one accessible environment. But just because we don’t have complete data sets, does that mean we should stop trying to identify audiences? No, not at all, it just means that we need to be more thoughtful about how we use data.

Imagine a data set that told you the following:


While these two audiences look different in comparison, what we don’t know is how different they look from the average. With analysis, the data set could tell us what the average respondent in the data set looks like from a demographic perspective and we could do a comparison that would add context to the information regarding diet versus regular soda drinkers.

This would provide the proper context to analyze the data, right? WRONG.


What if the original data set wasn’t representative of the U.S. population as displayed below:

Without questioning what the data set represents, you could make targeting decisions for your business that could have potentially disastrous consequences. Would you believe that both diet soda and regular soda drinkers are female and African American skewing?

While clearly this is a dramatization and smart clients have a handle on their consumers’ demographics, it is not hard to see how starting with faulty data can lead to bad business choices. There is a place for big data, but if you are using data (big or small) to identify audiences, you need to ask your data providers where it fits into the larger puzzle, to ensure you are making the smartest possible business decisions. Remember, friends don’t let friends get drunk on big data and make bad business decisions.

About Annalect Audience Exploration

We help to identify your most valuable consumers and recommend the best way to connect with them.

Leave a Reply

Your email address will not be published. Required fields are marked *