Analytical CRM
For many companies the first step in managing their sales operations is to introduce a Customer Relationship Management application (CRM) to optimise their customer relationships and effectively monitor their sales operations. However, for a CRM strategy to work it is first necessary to thoroughly analyse and understand customer behaviour. This is precisely the kind of task for which STATISTICA analytical CRM solutions are designed. Using data mining techniques, STATISTICA analytical CRM solutions analyse customer relationships and allow marketers to classify groups of customers according to their characteristics and buying behaviour (segmentation). They also help to improve the effectiveness of marketing campaigns and attract new customers, as well as maximising the value of sales to existing customers (cross-selling and up-selling) and minimising customer loss (churn). Data mining techniques are also used to analyse and monitor levels of customer satisfaction and loyalty and diagnose the causes of changes in these levels.

Segmentation
To succeed in a modern economy a company must accurately identify potential customers for particular products and services and formulate their offers to address the individual needs of their target audience. One way of achieving this is by market segmentation. It is impossible for large companies operating in mass markets to establish the preferences of each customer individually. They must therefore rely on data analysis techniques involving market segmentation to divide customers into groups of similar individuals and then select an appropriate marketing approach for each group. Segmentation also allows coherent and precise definition of customer groups and a better understanding of their behaviour and motivation. The information this generates facilitates selection of target markets, adjustment of pricing policy, selection of appropriate distribution channels, and more effective identification of competitors.

Cross-Selling and Up-Selling
Modern marketing involves not only securing new customers but also building lasting relationships with existing customers and maximising the benefits resulting from these relationships, for example by cross-selling and up-selling. A company should understand customers' needs and behaviour patterns. Since information about customers and their actions can often only be found in huge datasets, the best way to achieve this understanding is by employing modern data mining techniques. These allow the extraction of new dependencies and customer behaviour from data, which can then be presented in the form of logical rules for the most frequent buying patterns. Having established correlations for a particular group of customers, managers can then put together a suitable offer for customers with similar features or purchasing preferences, and attract their attention by creating new products or expanding existing product lines. Data mining techniques are also used to create predictive models for identifying target customers for a particular offer.

Customer Loyalty and Migration Analysis
In competitive markets with a low degree of product differentiation, where the cost to customers of switching to another supplier is low, customer loyalty is an issue of central importance. The ability to accurately predict migratory phenomena is vital for companies in this type of market. It allows them to take preventative measures, forecast sales figures, assess the total value of the customer (i.e. expected income generated by a customer over the entire period of his relationship with the company) or modify their offers. Data analysis methods, especially data mining, can be used in a number of ways to prevent customer migration. In particular, statistical techniques are used to analyse customer satisfaction, identify factors determining customer satisfaction and loyalty, monitor changes in the level of customer satisfaction, and identify customer behaviour patterns, thus allowing a company to adapt its offer to suit their needs. One of the most common approaches is the creation of a data-mining model to predict the probability of customer defection (churn).