Currently every business, regardless of size, uses some sort of spreadsheet technology like Microsoft Excel, Apple Numbers or Google Docs Spreadsheets to analyze, review and report business operations and performance. They create tons of multiple versions of the same documents and the most sophisticated users generate spreadsheets using data from different applications, such as combining information from ERP and CRM systems to conduct analysis and draw conclusions.
This is a daunting task that is prone to errors due to its manual nature. Unfortunately, businesses seldom have a better choice or motivation to switch away from the way they have been running operations for years, which is particularly the case with small and medium businesses. There are several problems associated with running your business on spreadsheets. They are hard to maintain, tedious to update and their intended result can shift, losing focus of the original purpose for generating such representations of business data.
Businesses spend most of their time gathering data points, matching up information from different sources, cleansing the data from what you do not need, joining pieces and trying to understand how applications process this information to help drive decisions. A large amount of time is also spent fixing mistakes and validating what has been built. This leaves little time to actually conduct the necessary analyses and apply insights to actual business performance.
The common IT solution for “spreadsheet hell” is the use of data marts. A data mart is an access layer sitting on top of the data warehouse the gives data access to users so they can resolve most of the issues they experience with spreadsheets. Getting a data mart is not simple and requires a series of technical components such as connectivity, access, extraction, transformation, loading, cleansing, mapping and data aggregation at the source applications.
By using data marts, business users do not need to do any of the tasks involved in terms of maintenance and data massaging. Technology experts pre-develop processes to access and extract information from applications. Data architects then transform the data giving it business friendly meaning. The result is that any business user could use and understand the data. This proven IT process also verifies that the data matches and maps to other applications. Finally in data marts, only the most meaningful metrics are aggregated like amount, quantity, unit cost, hours and cost, enabling a better sense of the business.
Data marts do all the work so that business can focus more on analysis instead of dealing with the raw data. Reducing data silos, human error, and most of all the tedious, time-consuming effort is essential to making important business decisions: decisions that impact the bottom line. However, data marts are just one part of the puzzle.
In the next segment we will describe the limitations of data marts and how analytics marts and dash marts will be the next evolution in the progression of business intelligence.