Understanding the Difference Between Correlation and Causation in Marketing

– Many marketers confuse correlation with causation when analyzing data.
– The fact that people mention TikTok as a source of discovery does not necessarily mean that investing more in TikTok will lead to more sales.
– Attribution in marketing is complex and often influenced by variables that cannot be controlled, such as timing.Question the relationship between data and marketing goals, and focus on understanding the impact of data on the bottom line or customer behavior rather than relying on correlations or assumptions.Most marketers don’t understand the difference between correlation and causation. Here’s what I mean: Post-purchase survey question: How did you hear about us? People are telling us TikTok! The data says that people are hearing about us on TikTok, but it’s not showing in our attribution. Does that mean we should invest more in TikTok? Marketer: Yes, we should definitely. People are telling us that’s where they found us. The real answer: No, not necessarily. What it means is a portion of people who are purchasing from you, and an even smaller portion of those purchasers who are filling out your post-purchase survey, are more likely to remember you from their browsing on TikTok. The marketer’s brain says, “We have a connection!” No. You’re confusing both correlation and causation. It gets worse. The only correlation you can determine based on evidence is: Users of TikTok who also purchase are more likely to fill out your surveys during the post-purchase process and attribute their discovery to where they already spend time. Well, that’s not as much fun as people buying because of our TikTok. The problem is marketing doesn’t go well with logical reasoning. It’s a problem across our entire industry. We look for stats and figures to support our goals without questioning their relationships. A true data-focused company needs to be able to pull a relationship to the data collected with factual evidence of how it impacts the bottom line or behavior. This is damn near impossible to do because of the number of variables involved. So we settle instead for soundbites that make us “feel” like we’re making a difference. Attribution is a collection of events that happen with a customer, with one that you absolutely can’t control: TIMING. Helpful reminder: Most companies can’t accurately calculate the lagging indicators of purchase because they don’t accurately track from awareness v. intent. No one can tell you when someone will purchase or when they are likely to purchase with any specificity for the vast majority of products. It’s all a confidence interval. So for all these agencies that only focus on one part of the business, all these companies that only focus on one part of data collection, all these experts that keep promoting data this and data that, stop kidding yourselves. The truth is, while everyone is selling the idea of learning more about your customers, most of the consultants, apps, etc., can’t tell you what intent data actually is. Most just install a tracker and use machine learning to find a trend between behaviors. If they could succeed with all the claims they are making everyone would use them and everyone would have success, we know that’s not the case. Start questioning how things fit together. #marketing #ecommerce #strategyhttps://www.linkedin.com/in/jivanco

Leave a Comment

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

Scroll to Top