We know that customer retention and loyalty management is crucial to long-term business success, especially in a slow-growth economy. Incentive programs can be a great tool to drive loyalty but with endless incentive design options, the best way to ensure that your investment in an incentive program will yield the maximum ROI can be summed up in three simple words—know your customer.

The issue years ago was getting relevant information about customers; today the challenge is that companies are inundated with information and it can be difficult to figure out what can and should be used to support the business. One company we worked with approached us to help design a program that would incent their members to leverage pre-negotiated contracts with select suppliers. This company’s past efforts in trying to incent this behavior through relatively generic, stereotypical programs were unsuccessful because, simply put, they were not targeted enough.

In this case, when we asked the executives questions about who their customers were (e.g., about what percentage of your customers earn over $100,000 per year) they answered confidently and were sure that they had a good handle on these statistics. But, did they really?

We ran their 100,000+ members through a rigorous analysis, pulling information such as income levels, charitable donations, number of children, and residential addresses. The results were surprising to the senior staff who thought they knew their customers. The real data opened their eyes to challenges in compliance and provided this client with new opportunities to drive the desired business results.

For example, through our analysis we found that a high percentage of the people we analyzed contributed to a certain charity, had 1.5 children and lived in rural areas. Based on this information, we suggested developing a distinct segment and creating an incentive program for this that would match-up with their particular group.

Depending on the industry and client, the amount of information available and that which can be used under regulatory guidelines varies. However, we have found that it is possible to work within these confines and still be able to use information to better understand clients’ needs.

Retail banking is one industry that’s already discovered the value of customer analytics and this trend is poised to continue. In addition to loyalty programs, particularly for cards, the study of consumer behavior can help build the bundled service packages that drive both revenue and customer satisfaction for banks.

Banks are now paying particularly close attention to their high income/high net-worth customers and how this group can be better incented, more efficiently serviced and provided a top class customer experience. This group may warrant a significantly altered customer service model, focusing on how the bank sells, communicates, incents behavior, services and markets products. For example, since a high percentage of these customers use iPhones, the bank might develop a mobile app that allows them to communicate and bank through the bank’s iPhone app, and receive and redeem incentives in real time.

However, there is a point at which businesses can over consume data. For example, when you start defining dozens and dozens of customer groups, the distinctions become too fine to be of practical use. Worse than being superfluous, they can cause a business to become paralyzed by the data and mired in indecision. In my experience, the best way to combat TMI is to think strategically when segmenting data fields and identify attributes most closely tied to what you’re trying to accomplish.

Based on the desired outcome, it is first critical to understand the attributes that influence those decisions. Then using these criteria, push the customer data through the system to create the desired segmentation models. If you’ve adapted an incentive program for a specific customer group, is the new incentive program working? If it is not, you will need to adjust it on an ongoing basis. No analytics program should begin without asking what the measures of success will be. When tracking trends, there will always be outliers or exceptions so you must have measurements in place. AB testing (in which a control sample is compared to single-variable test samples), using a combination of multiple attributes, can help you tweak your strategy to achieve success.

One of the biggest mistakes I’ve seen businesses make with a customer analytics initiative is treating it like a project. Customer analytics requires ongoing commitment and dedication to leverage the data to achieve business results. The data set is a tool and you will need to experiment with ways to make it effect the change you want.

Whether due to economic conditions that affect spend patterns, or the naturally changing demographics of your market, your customer base is changing over time. Are you on top of these changes? And are you ensuring that these changes are driving the required ongoing monitoring and maintenance of your loyalty programs?