3. Understand the underlying data issues
It is best to understand the data logic at the planning stage. Review the data needed from a business user perspective and then define the data relationships among the data. Design a data structure of dimension tables if possible where data elements for filtering, lookup, and sorting typically are defined and do not change significantly from week to week.
Then, create fact tables where measures and calculations are defined. Also, create a star schema to define how the dimension table and fact table interact. This sets the rules as the results are dimensionally conformed tables that all data will conform to.
This initial work of creating inter-related tables may be tedious at first, but they define the data structures and set the rules for how the data connect. Moving own, project teams will rarely have to make major adjustments, or worry about shifting variables and data structures.
4. Select the appropriate toolset -- look for a fast, easy and lightweight application that is affordable
Speed is usually an important consideration for data analysis and decision making. In the business world, it is about time to market and flexibility in reacting to market changes. A solution that is easy to use, and requires little to no training would be ideal in a user-driven business intelligence solution.
For example, Singapore based fashion company Lift12 utilizes data gathered from social media and purchase patterns to understand customer preferences. They then constantly update their designs and release new designs every few months based on the information.
Lift12 needs data analysis to be done quickly and accurately. Tapping on a self-service data analytics software, Lift12 is able to easily integrate, analyze and interpret over 17 gigabytes of data from different sources to produce reports on market trends. Data analysis work that used to take two weeks now requires only 30 minutes - time saved is time gained in the fast paced fashion industry.
5. Build for flexibility and responsiveness -- know that users will require changes early on
Business intelligence users will need to change their requirements as business grows and evolve to adapt to market changes. Users are bound to have additional requests along the way no matter how the team built the initial workbooks and reports.
It is important that teams design data structures to be comprehensive, and the business intelligence tool selected is fast, flexible and can quickly be modified to accommodate users' requests.
6. Leverage existing staff -- eliminate solutions that require new staff and/or consultants
Staff that already work with the data normally know their data best. They understand how the data relates to everyday operations, and they know what information users want to get out of it.
Sign up for CIO Asia eNewsletters.