Using Correlation Analysis for Customer Experience

By Roy Barnes, Bob Kelleher

Naturally, you, your chief financial officer (CFO), and your chief executive officer (CEO) are probably wondering just how much you have to improve your customer experience score to move the three key metrics — recommend, switch, and repurchase. To figure this out, you need to perform correlation analysis.

Correlation is simply the relationship or connection between two things. For some, chocolate and a happy life are highly correlated. The more chocolate they eat, the happier they are. Mushrooms and a happy life are also highly correlated for some people, but in an inverse manner. The more mushrooms they are forced to eat (or look at or be in the same room as), the unhappier they become.

Both of these examples have what data geeks call a “strong correlation.”

When correlations are calculated mathematically, the result is a correlation coefficient. A correlation coefficient is a number between ‒1 and +1 that measures the degree of association, connectedness, or linkage between the two variables. Correlations above 0.7 or below ‒0.7 are considered to be strong.

In recent years, researchers have proven a strong correlation between various elements of customer experience and the three key metrics — recommend, switch, and repurchase. While correlation does not mean causation, it does mean that with robust data about your customer’s experience, you can begin to make logical connections between experience and financial results.

The question becomes, which activities and initiatives will have the biggest impact on customer experience — and by extension, on the aforementioned key metrics? To answer this, you must test, test, and test some more. Start by looking at all the different correlations between your various experience metrics and financial metrics. Which experiences move which financial metrics?

During this process, you want to meet with the strongest analytics people in your organization. Ask them questions like, “What is the repurchase behavior for customers who score an 85 percent in overall customer experience/satisfaction compared to those who score an 86 percent in overall customer experience/satisfaction?”

Answering this question enables you to calculate the financial payoff (in terms of repurchase) for each percentage point of improvement in overall customer experience. This should in turn help you determine which customer experience initiatives you may want to tackle first.

For example, suppose you have two different initiatives, each of which costs the same amount of money, but you can fund only one of them at a time. Initiative A may lead to a three-point increase in customer experience, but Initiative B will lead to a five-point increase. In this scenario, pursue Initiative B and execute like crazy in the hopes you’ll see the additional boost.

Here’s another question you should ask: If you can move the overall customer engagement score by 1, 2, 5, or 10 percent, what happens to customers’ price sensitivity? If you can prove that happier, more engaged customers are less sensitive to price increases (which they often are), your CFO probably will take you out to lunch. And maybe even pay.

When you get in the habit of probing for the correlations across various metrics, you begin to legitimize your organization’s investment in customer experience. Be curious. Ask “dumb” questions; don’t be afraid to test all sorts of variables to see which ones yield the best results. What you’re looking to find out is what happens to each of your financial metrics when you modify their inputs.

For an example, see the following figure. Here, it’s been determined that touchpoints #1, #3, and #5 are the most influential in driving a customer’s intent to repurchase. (A touchpoint is any point where the customer and company interact.)

[Credit: Illustration courtesy of Roy Barnes.]

Credit: Illustration courtesy of Roy Barnes.

If the organization wants to improve this repurchase metric, then it would likely be smart to focus its redesign efforts on these most impactful touchpoints, placing less impactful touchpoints lower down the list. Similar analysis could be conducted to determine which touchpoints are most influential in the “intent to refer” and “reduce likelihood to switch” metrics.