This site uses different types of cookies, including analytics and functional cookies (its own and from other sites). To change your cookie settings or find out more, click here. If you continue browsing our website, you accept these cookies.
Why play fair when you can stack the deck in your favor?
Whether you are selecting down to a smaller group or simply sampling from a larger audience, why not be choosy and pick the best data for your next marketing effort?
1. Empirical Evidence
In science, empirical evidence is required to gain acceptance of your hypothesis. In business there are lots of decisions based upon tribal knowledge and gut instincts. Data exploration can work with small quantities of data and produce big paybacks. Blend your data with 3rd party data and create new insights that reveal new truths.
Be picky and choose wisely. When everything is equal, keep it simple and choose randomly. But in most cases, random selection should be your last choice. If you’ve got to choose which individuals or households to spend your marketing dollars on, you must spend it first where the greatest response is forecasted. Use your empirical evidence and move from good to better through profiling and segmentation. Move from better to best by modelling your data and use multivariate analysis.
3. Using Random Wisely
When you do use random selection, please test your results and make sure that the selected records represent the same data distribution as the parent group. If you are going to make big decisions from small numbers, you really want accuracy in your data. Small mistakes could lead to big conversations later. In testing, you will ultimately need a random function. As a best practice consider using a Chi-Square test (for categorical data) to make sure that you’ve got a good sample.
If you’re an Alteryx user, you can download a Chi-Square test macro from the Alteryx gallery that simplifies the check for a representative sample: https://goo.gl/YfgrOj. I've loaded the macro into the gallery for you to use.
It’s always a good idea to list your assumptions and test them with the real world. See where the data takes you and master the obvious while you challenge yourself to become a myth-buster.