Data Warehousing EAIs Education and Career Information Security IT Project Management IT ROI
Outsourcing and Offshoring Post-merger IT Integration Public Sector IT Consulting SAP Website Analysis

Data warehousing

The larger a business is, the more data it generates. Up to a certain point, data can be stored without much of a strategy, but that point comes sooner than many companies expect. If your data storing solution works fine, it still doesn't mean that the data is used optimally. To have business value, it has to evolve into knowledge. That’s when expressions such as data warehouses, data mining, data management, decision support and business intelligence come into light. Some of them have overlapping or even disputed meanings, but they all boil down to extracting information from past data and using it to make better business decisions and thus gain competitive advantage (if you don’t mind the overuse of this phrase).


Analyse this

A data warehouse is usually a relational database that takes data from many systems across the company. The data has to be analysed to find patterns and relationships that can predict the behaviour of customers, prospects, and even the enterprise itself. In a retail chain, for instance, thousands of products are sold and kept on stock. Seasonal sale patterns can be used to reduce stock – and expenses as well. Daily and weekly sale patterns could lead to optimising deployment of staff in the stores, and so on.
Note this “lead to” element – going from raw data to knowledge and from knowledge to decisions that actually benefit business is a challenge that can’t be overcome without expertise. There aren’t generic data warehousing solutions – data is produced differently from one company to another, and the information that is crucial to one business can be useless to the next.


Data warehousing is not for everyone

Most articles that present data warehousing do it from an affirmative point of view and give examples and arguments for (more efficient) data storage. Not a lot is said against it, although there are quite a few things that can recommend against constructing a data warehouse.
First, the business can simply be too simple. It’s not about the quantity of data, it’s about whether it contains information that isn’t readily accessible or obvious without data warehousing. Second, even if the IT part of the work is completed quickly, it takes time for an organisation to find out how it can change its business practices to get a substantial return on its data warehousing investment.
Third, a data warehouse by no means runs on its own. New data is added all the time, end users keep coming up with new requirements – the more the data warehouse is used, the more staff it needs. And that means high maintenance costs. Sometimes capturing data, cleaning it up, and delivering it in a format and time frame that is useful for the end users, simply doesn’t pay off.