"4 out of 5 Dentists Recommend.....Really?"
Working in the data industry is certainly fascinating. It amazes me just how much data is available to marketers in order to find their niche prospects, or to better understand their current clients.
For example, if you wanted to find dog lovers that live in a certain zip code with a yard size of an acre or more and a household income of $75k or more...that's a piece of cake! What about divorced sales people with an interest in religion and political donors? Sure! Maybe you're a bank wanting to offer credit cards to people located within five miles of their branch locations within a specific credit score range that have a home valued at more than $500k and...that moved from out of state in the past year? Come on, that's too simple.
With thousands of data elements from demographics to lifestyles, to interests, to buying habits, there's almost an endless amount of information that is being used to improve marketing, develop products and services, and many other things. It's truly not a bad thing in most cases as companies use this insight to make improvements on what they market to you and how. After all, wouldn't you like every piece of ad mail or digital ads to be pertinent to you? What isn't necessarily "fun" but is still very interesting is the science behind the data and how...when done properly...it is very insightful, BUT how it also causes me to question every poll, survey, or study that I see.
Having worked with our Data Analytic Team on many and various projects, it is critical to build your analytical work with a strong foundation of accurate data and sample selection. Whether building a customer profile, clone model, or predictive model, it's very important to find a truly accurate representation of the general population in which to compare. Also, having the confidence that the data elements that are used to gain insight into your customer base is equally accurate.
Much like a survey, poll, or study, if the sample size is too small, or biased (intentionally or not), or any other part of the sample selection is corrupted (again, intentionally or not) the accuracy and credibility is destroyed. As a Marketer, it's fair to ask the question of how your Data Scientist who is developing your customer profile, or model, or whatever the project may be, "how are you developing your sample?" After all, you don't want a predictive model that predicts the wrong outcome.
Data Science, specifically in the area of developing marketing insight, can be EXTREMELY helpful in developing your target market, but don't hesitate to ask "how and why" early in the process in order to make certain to build your analysis on the best foundation possible. Also, don't blindly accept everything you hear or read in regard to polls, surveys, or studies. Look for the facts. Consider the possible bias. Be a critical thinker! After all, the "4 out of 5 Dentist" that recommended Trident chewing gum may have been family members/friends of Trident employees, or paid to endorse, or stockholders in the company, or.....you get the point.
If you'd like to speak with one of our Data Analytic Team members or receive more information by email, just click below and complete your information. We're here to help!
Don't miss out. Catch up on our blog posts at blog.altairci.com. Subscribe today!
Many counts can be found with our online system at Altair Online: http://counts.altairci.com/