After more than 15 years developing analytic solutions to enable companies to understand and act on customer behavior, I still see businesses of all sizes struggle with how they can use an understanding of their customers to accelerate their business.

Despite the buzz around “customer-centricity” and “big data,” the reality is that most companies do not take full advantage of everything they know about their customers or their relationship with them. The questions are typically, “Where do I start?” or, “How much will it cost?”

While the scale and resources of a large company can be advantageous, the idea that only giants can leverage data analytics and draw on actionable insights from their customers to grow their business is false. While outsourcing your customer insights is an option, it is not always a best practice. Every company should understand the analysis rather than be completely divorced from it. So whether you have the resources to tackle customer development and data analysis on your own or you prefer to work with a partner, here are five basic principles for any business to ensure it will be effective in achieving your business goals.

1. Put customers at the center

All businesses interact with customers. But we often look at these interactions in aggregate instead of by type of customer. A good place to start for any organization is to take their existing data and look for ways to pivot it to be around the customer. For example, you may know your sales, but do you know your sales data for just first time buyers? What are they buying? Or you may know your mobile app traffic, but how much is from men versus women? The first key step in this process is to also organize your data by customer instead of just by product line, channel, brand, or store. This customer-centric approach will build the foundation for more sophisticated analysis in the future like predictive modeling.

2. Link or derive new data for more power

The more you know about your customers, the more powerful the data becomes. There are many places you interact with your customers; have you linked all of that data together? Is your sales data tied to your call center data? Do you know who likes to shop during certain types of promotions? For on-line retailers, have you linked your site browse and search data to your sales data?

In addition to linking data, a powerful approach is to label customers with derived attributes. For example, who are your “best” customers? Which categories do they typically buy in? When is the last time they shopped? Once you have started this foundational approach, it is easy to add additional “genes” to this Customer DNA.


Data iconRelated research on VB Insight:
Neal Ungerleider’s report Big Data — Meet Dumb Data


3. Ask your customers

It’s important to realize not everything you may want to know about your customers exists in the data you collect. It’s critical to match the quantitative information you have from doing business with the insights you can gain by simply talking to your customers, which may be in the form of survey responses or even free-form comments. Asking your customers questions and soliciting feedback, then combining the responses with the data-driven insights you already have, enables you to develop a more informed picture of what actions to take.

Take housing as an analogy. A broker only looking at the total number of completed mortgage applications is missing out on a host of insights. By surveying individuals who started, but failed to complete applications, he or she would be able to understand what is holding potential customers back – whether it’s fees or the complexity of the forms – and identify ways to improve the process.

4. All customers are not created equal

If you are only solving for the average customer, you are not really solving for anyone. Think about how customers are different from each other, which ones you care about, and how you are going to tailor your actions for the groups you want to focus on. What inventory will you carry? What marketing promotions will you do? Don’t view variability among customers as a complexity, view it as an opportunity!

For small businesses, this does not have to be a daunting task. As an example, it can be as simple as tracking what people buy from you on an individual basis and breaking that information down into a few buckets. This is basic customer segmentation, but even having an understanding of your high, medium and low value customers allows you to identify and tailor actions around how best to engage with each group or individual.

5. You can start small, but make it repeatable

It’s easy to drown in data. The key is to take one specific process or customer touchpoint, make changes based on data for that specific purpose, and do it in a way that’s repeatable. You don’t need an army of analysts to be successful. The key is that without a process — whatever the data set size — you won’t succeed.

Take a single interaction point and go end-to-end – from collection, organization, to analysis, to implementation. For instance, determine who your high value and low value customers are. Based on these two groups, personalize some of your communications to these groups, whether that’s promotions or messages on the receipts you send. By improving just one type of interaction, you can begin to scale it to other parts of your business.

Zoher Karu is Vice President of Global Customer Optimization and Data at eBay, where he works to use data, analytics, and insights to drive growth across all customer interactions, with both buyers and sellers. Prior to joining eBay, Zoher served as Vice President of Marketing Analytics and Insight at Sears Holdings, leading efforts to drive customer behavior and loyalty, both in-store and online.