A rip-off from Gartner and Kimball but what the heck,

The initial build, is the first raw version of the data warehouse or data mart for that matter. Usually triggered by some consultancy moving in and developing a quick and simple DW. Often together with influential and knowledgable staff in the organisation. Low cost and fast results is key.

The fast track to success, is the phase when more data and new features is added to the data warehouse, still driven by the ”founders” of the system the sky is the limit. Problems are still fairly simple to solve and business benefits are high for each new investment. Key individuals connected to the data warehouse are given status as ”gurus” or ”doers”. On the horison the dark clouds are looming.

The quality crisis, at some point in time the complexity of the data warehouse have reached a level where the initial fast track solution becomes hard to manage. Data sourcing, policies and applications diverge and raise concern regarding how good a data warehouse it actually is. In this phase it’s common to find anomalies in original statistics, such as errors in calculations, hidden errors and other problems that in worst case affect external image of a company.

The painful slope of new confidence, after the quality crisis the cumbersome work with restoring good faith in data begins. Data is better structured, better controlled, better understood and usually the development and maintenance is better organised. In the end the crisis increase quality and give the organisation a better grip of it’s operation. The pace of application development slows down in order to keep a high standard of quality.

The monolith phase, the data warehouse finds itself ageing and might well be challenged by new data warehouses that are emerging beeing earlier in the lifecycle. Old quality problems might still trigger questions among users and reputation might still be bad even though the data warehouse now are delivering good results. Other problems occur, usually related to new needs and new influentials that are feeling their needs are not accomodated in time.

The consolidation phase, in a major organisation possibly where there have been a few mergers and aquisitions the portfolio of data warehouses, data marts and analytical applications. The CIO might at this stage feel compelled to initiate clean up operations in order to reduce IT-costs and complexity. So the initial high profile data warehouse might find itself sucked into a black hole, ”the enterprise data warehouse” or left to die a slow death of continuously reduced budgets.