Causative vs Correlative – Understanding your Data

In marketing, data is everything.
If you understand how to leverage your data and analytics, you can more effectively reach your target audience, create more engaging content, and ultimately increase your ROI.
But, how do you distinguish between real insights and red herrings? 
This starts with understanding the difference between causative and correlative relationships.


Let me give you an example — a friend recently brought up this fact:

Males who regularly use the sauna (4 – 7 times per week) have a 40% reduction in risk of cardiovascular events and all-cause mortality. [Source, 2015]

You may be thinking, “I’m getting in my car right after reading this article and driving to the sauna!”
But you need to take a step back and think if this relationship between saunas and mortality is causative or correlative.  Are saunas really the reason for a reduced risk in cardiac events or is there just a correlation here?

  1. Sauna users typically live a healthier lifestyle.  Most people don’t have a gym membership just to use a sauna.  Most sauna users work out before getting in.
  2. Sauna users typically come from a more privileged lifestyle.  If you can afford access to a sauna, chances are you have healthy living conditions, and you can afford quality food.

So there is a correlation between saunas and mortality, but it is unlikely that saunas are the cause for a reduced risk in cardiac failure.

How it Applies to Marketing

You need to establish the same holistic viewpoint when analyzing your marketing data.
Someone might argue, brands with revenue north of $10 million have over 100 thousand followers on social media.
That may be true, but don’t think that if you go and gather 100k followers suddenly you’ll be making $10M ARR.  These companies have built strong brands with a solid product and many customers.  The social media followers came after.
Another example would be social media accounts that post 10+ times a day have millions of followers.  Take BuzzFeed for example — they have 6.6M followers and tweeted 260 times in the last 7 days [source].
Screen Shot 2018-05-31 at 3.20.24 PM.png
If you only had 100 followers and tweeted 260 times, you would quickly annoy those 100 followers and begin to lose a following.
The point being, ensure you understand the whole picture before making any sort of assumption.  Look at multiple sources and find a thorough reporting tool such as Cloud Campaign.

Leave a Reply

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