When it comes to customer information, these are the best and worst of times for corporations. The digital era has made available rich new sources of data about customers. Yet rapid growth in customer segments, distribution channels, store formats, and product categories means companies must combine and integrate that information in newly complicated ways.

Specifically, marketers must look at the intersection of different types of information (for instance, between customer needs, store formats, and product types) if they are to convert mere data into competitively useful insights about customers—that is, into knowledge beyond conventional industry understanding about why, when, how, where, and what customers buy. Useful insights are the product of two or more combined pieces of information. The common characteristics connecting seemingly dissimilar groupings of customers only become visible when viewed from multiple angles. By focusing on these intersection points, or "cells," marketers can better avoid averaging out consumer preferences that, if properly understood, could suggest new opportunities for growth.

Most companies, though, still regard customer insights as an isolated research capability. As a result, they aren't configured to obtain data at the points where segments, channels, and categories intersect, nor can they integrate the information to generate valuable insights. The isolation of the insights function also inhibits transforming insights into actions and leaves many companies without a common way of looking at and describing customers across functions. Instead, marketing focuses on brands; sales looks at geography, channel types, and key accounts; and the market research organization views the world in segments.

Get the full story at The McKinsey Quarterly (free registration)