Data is a great resource for the business owner who wants to better understand their customer and make better decisions. When we examine data, we’re ultimately looking for explanations – and this is where many business owners run into trouble.
Many times, we think we’ve arrived at understanding long before we actually have. A quick glance at the numbers, trends and patterns isn’t enough to help you. It’s important to invest the time to understand what factors have contributed to the outcomes you’re examining, and whether the relationship of those factors to the outcome are correlation, causation, or coincidence.
Exactly what is going on here? Examining the data for causality
Let’s say you are a specialty retailer – a vinyl record dealer – who wants to understand a dramatic but short-lived spike in their website and social media traffic, as well as in-store sales. Knowing what caused the spike might allow the retailer the opportunity to recreate those conditions that drove customer demand – obviously a desirable outcome.
It’s important to remember that data gives us a picture of what already happened. If we want to understand the activity spike, we need to look at the data captured when it happened, as well as the time period immediately prior.
Perhaps, as the business owner, you’re wondering if the sales spike might be related to the extra time and effort you’ve been putting into social media. Examining engagement rates does indicate that you’re connecting with more customers more effectively – but does this explain the activity spike?
The answer is we don’t know. These two data points – the engagement rate and the activity spike – may appear to be related, but we don’t know for sure. We have no idea if the engaged customers are the same people who are coming to your store to buy records. Sometimes people talk on social media because they enjoy talking on social media. There’s no definitive relationship between the two data points, even though both have increased. This is called correlation.
When we see correlation, we know that there may be something going on that’s worth paying attention to. Knowing that increased effort lead to increased engagement is encouraging; having understood that data point, it’s time to consider how to capitalize on that engagement effectively.
But for right now, we’re interested in causation – what factors definitively led to the activity spike. Let’s suppose the vinyl record dealer used email marketing to reach their customers. If an email message that invited customers to visit the website & use a specific savings code was followed by the activity spike, we have what is called causation – a direct, definitive relationship between your action and your customers’ response.
Once causation has been identified, it can be repeated. In this instance, investing further time and energy into additional similar email campaigns would appear to be a wise decision.
What if you can’t find causation?
But let’s say there was no email marketing campaign going on, nor any other causal element to be discovered in the data. What does the business owner do then?
It’s important to remember that customer behavior happens in context of a much larger world. We’re never going to be able to track all of the factors that could motivate purchasing behavior, but it’s a really good exercise to try. Reviewing the activity spike with the vinyl store owner revealed that a popular musician passed away immediately before the activity spike – and delving deeper into the data showed that the majority of sales during the spike were that musician’s work. This is a coincidence – a factor entirely outside of the business owner’s control still had an impact.
We can’t control for coincidence, although we can certainly profit from it when the opportunity arises. And it’s important to understand that we operate in the real world, where the boundaries of what does and doesn’t influence customer behavior aren’t as cut and dried as we might like. Did the vinyl record dealer discuss the recently departed musician on social media? Perhaps – and that may have contributed to the activity spike. Had there been an email campaign, would some of those customers have used their savings code to save on the music they were going to buy in tribute? Perhaps.
Marketing is not an exact science. We must do the best we can to remove the emotion from the analysis and avoid rushing to judgement. That being said, every day customers create more data points for us to examine, allowing for ongoing improvement in understanding and growth – if we’re willing to invest the time in understanding what we’re looking at.