Business Intelligence (BI) is a term given to the gathering, storing and analysis of data to aid users in making educated decisions. To keep track of the vast amounts of information an organisation collects/generates, it is often required to use a wide range of software applications and databases throughout their organisation.
Using numerous software packages makes it increasingly difficult to retrieve this information let alone format, aggregate and process the data. This is where Business Intelligence strategies help extract the optimum information for the business.
BI represents the tools and systems that play a key role in the strategic planning process of the corporation. They allow end users to sieve through large quantities of data, both internal and external to the company and aid in the development of business strategies.
BI systems are often used in the areas of customer profiling, customer support, market research, market segmentation, product profitability, statistical analysis, and inventory and distribution analysis.
Customer Relationship Management provides an integrated approach to identifying, acquiring, and retaining customers. Modern communications necessitate the need for a customer to be contacted via multiple channels to provide targetted products and services, and to improve customer/organisation relationships.
By enabling organisations to manage and coordinate customer interactions across multiple channels, CRM helps maximise the value of every customer interaction by ensuring any feedback loops required are closed off.
The challenge with CRM, is to enable customers to do business with the organisation, at any time, through any channel, while at the same time making the customers feel that they are dealing with a single, unified organisation.
As an overall strategy CRM seeks to improve business performance by identifying the customers needs and behaviours, and then matching products and services to satisfy them.
CRM is effectively a marketing philosophy based on putting the customer first with technology as a driving force.
ETL, or Extract, Transform and Load, are the processes that enable companies to move data between multiple systems, often reformatting or cleansing it on the way. Dependant on the origins and evolution of a company, data is often stored is disparate systems and in differing formats, yet for a multitude of reasons requires consolidation.
There are software packages available commercially for performing ETL, which may or may function as an efficient solution depending on the software systems being interfaced. Often it is a process that is best constructed from scratch to ensure the resultant process is completely robust.
However ETL is performed it should ensure data quality and regularity is maintained, while also being fully automated and providing a notification system for any issues it encounters, or tasks it successfully performs.
A data warehouse is a central repository for all or the significant parts of the data that an organisations different business systems collect. There is a multitude of benefits to having a centralised view of data including provision of a customer level view, a history of the interfaced data systems, marketing, revenue assurance, transactional or legal triggers, provision of unified information back to source systems etc.
The vast amount of data are collected in larger organisations means great storage and processing power is required to make sense of it. Data warehouses size and processing capacity are often limited by budget. For this information to be rapidly accessed and analysed, it is particuarly important that any processes utilising it are efficient.
Data Marts are smaller departmental databases containing only information required by the specific group, are more financially viable and quicker to implement.
Customer / Market Segmentation
Customer segmentation is the process of clustering a customer base into subgroups of individuals that are similar in behaviour and demographics. On the basis of the segmented groups similarity, they are more likely to respond in a similar manner to a given marketing strategy.
Statistical models can assist in the generation of customer segmentation. These models can be as complex or simple as required by the business. Input variables often include age, gender, location, spend, product holdings or product usage.
Modelling and segmentation is often used to assist in customer retention, to increase usage, or to personally tailor communications for a customer.
The use of control groups is recommended to be able to more accurately gauge the effectiveness of campaigns in relation to segments.
Modelling & Mining
Data modelling is effectively the process of designing a data structure in a format that is functional, understandable and efficient for the resources that will be using it. It is an abstraction activity whereby what is a data structure created to be suitable to the source of the data is converted to design more useful and efficient for the modelled application. In the process modelling will merge structures from many data sources into a final homogenous structure, along the way analying and eliminating redundancies while juggling having a final design not having multiple points of update for one data change vs database efficiency.
As the volume of data increases, the capability to intuitively analyse it for effective usage decreases. It is often becomes necessary to have advanced mathmatical and statistical knowledge. There are many techniques that can be used to find useful patterns in your data such as neural networking, regression analysis or support vector networks.
Database marketing is a form of â€˜direct marketingâ€™ using databases of customers, or potential customers, to generate personalised communications in order to promote a product or service.
The main differences between â€˜directâ€™ and â€˜database marketingâ€™ are analysis and quantity of data. Database marketing can utilise statistical techniques to develop models of customer behaviour, which are then used to select customers (Customer Segmentation) for contact.
Due to the larger volumes of data, database marketers often utilise data warehouses to increasing the quantity of data, about customers, which in turn allows more accurate models to be built.
Database Marketing basically consists of two major tasks, Marketing Analysis and the subsequent implementation of the resulting strategies.
Marketing Analysis is iterative learning process, in which future campaigns are improved by learning from analysis of the previous ones.