BI Consulting for Data Management and Analytics

Risks From All Directions – How To Mitigate BI Project Risks

Effective planning and communication are key to managing risks in BI projects. Unexpected delays and costs have a way of moving projects from the success column to the failed column. Common sources of risks in BI projects include:

• Data quality problems
• Data semantics issues
• Data availability issues
• IT application specialist resource constraints
• Enterprise architecture shortcomings
• Scalability issues of BI platform and feeding systems
• Differing organizational objectives
• Differing availability of SMEs and departmental management
• Lack of user training and follow-through
• Lack of executive sponsorship, commitment and clarity



In many ways BI project risks are not much different than for other IT projects except that BI projects typically cut across both system and organizational boundaries. Coordination, communication and consensus become much more problematic and critical.

Develop of a clear strategy and tactical plan are essential, as well as good followup with communication and management.

Pollution in the Data Ecosystem Accumulates Up the Foodchain

The biggest surprises in many BI projects come from taking a comprehensive look at the data in an organization\’s databases. It is a rare (if non-existent) business that followed a data quality management program from the inception of its systems. Five commonly cited attributes of data quality are the following – any experienced IT person will see many opportunities for data quality problems in this list:

• Completeness
• Validity
• Consistency
• Timeliness
• Accuracy

Consider that the modern organization has multiple databases, with at least some overlapping data, coming from any number of sources, with differing validation rules. And consider that legacy databases, which still hold the bulk of the world\’s data, often used the same field for different purposes at different points in time.

Consider all these possible sources of errors, and roll them up into a consolidated whole, which must then be consistent when viewed from different dimensions, and it is not hard to see the potential for months of unplanned tail-chasing.

Successful projects plan for the existence of all these problems, and plan for searching out and remediating the \”unknown unknowns\” through the effective use of data quality tools and experienced analysis may be part of a Data Profiling phase of the project. This will identify and correct problems such as errors with data types, minimum and maximum values, valid list values, formatting, expected frequencies, relational dependencies and so on.

Vision, Strategy, Plan…Team – For BI Success

Curious about how to structure your BI project for success? Give us a call, we\’ve got loads of experience to help you achieve your goals.