Database Marketing employs statistical techniques to develop customer behaviour models, which are then used to select certain customers for communications.

To enable Database Marketing, the database will need to contain a variety of data including: name, address and contact details; permissions; transactional history; demographics as well as communication and response history.

In business-to-business (B2B) Database Marketing, the number of customers and prospects will be smaller than that of comparable business-to-consumer (B2C) companies. Also, because relationships with customers will often rely on intermediaries, e.g. salespeople, agents, and dealers, the number of transactions per customer may be statistically negligible. Another complication is that B2B Database Marketers may have many contacts for a single organisation, and determining which individual with whom to communicate through may be difficult.

However, increasingly online interactions with customers are providing B2B Database Marketers with a lower cost source of customer information.

In Database Marketing, companies often analyse the data with a view to segmenting customers based on their behaviour, needs, or attitudes. A common method of behavioural segmentation is 'RFM', in which customers are placed into segments based on the recency, frequency, and monetary value of past purchases.

Predictive models are often developed and used in Database Marketing, which forecast the propensity of customers to behave in certain ways, for example Database Marketers may build a model that ranks customers based on their likelihood to respond to a certain promotion. Statistical techniques that are commonly used for such models include logistic regression and neural networks.