Business Intelligence Specialist : Business Information Analyst : Project Management : Account Management
Successful BI Project

Successful BI Project

Foundation for a Successful BI Project

At the start of every new software project, you must focus on formalizing the vision and gathering all relevant requirements. This applies twice as hard for initiatives in the field of business intelligence - mainly because one of the biggest reasons for failing BI projects is the inadequate determination of requirements.
What is so special about the requirements for projects in the field of Business Intelligence? Well, reports, dashboards and real-time notifications have little to no value if they don't allow anyone in the organization to make a better decision. Ideally, all decisions support the organizational strategy, so an explicit link between BI initiatives and the strategy is crucial for success.

BI projects where this link is missing end with useful but worthless reports and dashboards. All participants must be confident that the right tools are used to monitor the effectiveness of the overall strategy and that there is a correct link with operational measurement data.
This need for strategic alignment means that BI projects must always be supported by C-level administrators. After all, it is BI projects that are expensive and complex and also cover multiple departments and hierarchies. Compliance with government regulations, data integration (both within and outside the organization) and the possible need for solutions that are diagnostic and / or predictive in nature, all play a role.

To properly understand the strategy-driven requirements of a BI project, an organization can choose multiple forms of approach and methodologies. Ideally, the strategy is already formalized, preferably on a proven framework for strategic analysis. The gold standard in this area is Porter's Five Forces Model, but there are countless alternatives available.

Once you have an idea of the organisation's strategy, you can ask useful questions about what decisions need to be made to implement this strategy as effectively as possible. As soon as you know which decisions need to be made and by whom, you are well on the way to understanding your BI requirements for the most part.

Measuring performance
This is a simple example of strategic alignment in practice. Suppose a production company has opted for the value 'operational excellence / low cost' (from the book The Discipline of Market Leaders). The result is that the most important decisions have to be made around cost savings in the supply chain, downtime and productivity up.
In this scenario, the company determines KPIs for inventory, cycle times of the production line, defects, machine uptime, and so on. On the other hand, a company that has chosen 'customer intimacy' as its most important value will be much more interested in opportunities to learn more from customers, want to become more closely involved with customers through co-creation and the value of the customer throughout want to maximize the entire duration through cross-selling, up-selling and additional services. This company determines its KPIs on the basis of factors such as repeated sales, promoter score, customer contacts and sales per customer.

A valid BI initiative also includes an analysis of how data and insights from the BI solution are used to validate the chosen strategy. This issue falls within the domain of performance management and organizational development. For example, a company may decide to use methods such as the balanced scorecard or BGC matrix to evaluate the strategy and determine the value of improved information provision.

Stakeholders
The next point in getting the right requirements for BI is addressing the question to what extent you have sufficiently involved all stakeholders. To get the maximum value from a BI project, you must identify all stakeholders and take every perspective into account in the following steps, regardless of where the stakeholders are located in the organization.

Let's go back to the example of the production company. The primary data that a foreman needs is, for example, maintenance time measurement data for a series of machines on the production line because their main goal in business strategy is to optimize productivity by reducing downtime. A middle manager in the same company may be more concerned about productivity on a per-factory basis. In this case, two employees have the same objective but a different data requirement. Developing a BI solution without carefully considering the needs of both stakeholders - in addition to those of others - would be a costly mistake that would reduce the value of the project.

Access control is also related to this issue. Many companies strongly restrict access to data and impose a too rigorous 'need to know' policy. While it is important to protect sensitive information, it is equally important to ensure that a BI system meets the needs of all users, regardless of rank in the organization. First-line employees, who are closest to problems when they arise, are often best placed to solve problems and make important decisions. Nobody gets along with it if you remember crucial data.

Compliance with laws and regulations is another concern. This is especially important in companies whose shares are traded publicly or in strictly regulated sectors such as finance, healthcare and government. For these organizations it is essential to ask for advice from the department that deals with compliance, the CFO, the legal assistant and other stakeholders who are charged with legislation and regulations in the portfolio. Anyone who does not do this can face fines, product recall and image damage.

Data Requirements
Finally, BI projects also depend on a careful approach to data integration. Data that is needed can be located in multiple operational data stores, in existing analytical databases or even outside the organization in the form of open data.
In the worst-case scenario, the company does not have "a single version of the truth" for important records. Many companies still work with a separate logistics database for customers in addition to the CRM. In these environments it is ideal to first implement Master Data Management and then set up the BI system, although this is not always practical.
In this case it is important that if the BI team has to get started, they have a complete understanding of existing data schemes and carefully manage the ETL flow to the data warehouse / data lake and ensure that reports and dashboards for all users in the organization are reliable. Nobody wants to get caught up in a situation in which multiple participants with different figures on the same subject come from different systems in the same meeting and have therefore drawn a completely different conclusion.

Key to success
BI projects tend to be larger in scope and costs than previously estimated. The requirements phase often already determines whether a BI project will help the organization to move forward or is simply a waste of time and money. These are the main requirements for collecting requirements within a BI project:
  • Ensure a clear business strategy and know how the BI initiative can support this strategy.
  • Know well which decisions must be taken based on the BI system.
  • Establish the needs of all stakeholders, regardless of their level in the organization.
  • Provide KPIs and measurement data that support decision making by stakeholders.
  • Note the overlap between BI and performance management.
  • Take account of legislation and regulations and ensure that the right stakeholders are involved in discussions about this.
  • Get to know the current data environment and its impact on the BI solution. If possible, implement MDM before you start working with BI. If that does not work, work carefully on the data quality of the data warehouse.
Addressing these issues when gathering requirements is not a guarantee of success, but if you do not look at this, I can guarantee that you will not get the most out of your investment. A well-designed BI solution can be invaluable for an organization, so make sure you don't go wrong with the requirements and build a system that supports the strategy throughout the organization and takes everyone's needs into account.

Share by: