Article by M.R.S. Company Ltd.
To have a successful Business Intelligence Deployment in today’s world, it must start with a comprehensive vision of the goal prior to any technological discussions. Here are 10 elements that you need to be on your way to a successful Business Intelligence Strategy
1. Choose a C-level sponsor (who is not the CIO)
A suitable sponsor for a BI program should be an executive who is responsible for the bottom-line, has a broad idea of enterprise objectives, strategy and goals, and knows how to translate the company mission into performance indicators that will support that project. The CFO is an ideal sponsor since the job often encompasses these three features.
2. Establish common definitions
Since BI relies on automation, it is imperative that each term is understood similarly between different employees and different departments. The relevant staff must work inter-departmentally to establish clear, accepted definitions. For example, “gross margin” may have a different meaning in the finance department compared to the understanding that the salespeople apply to the term.
3. Assess the current setting
Boris Evelson (a Forrester Research analyst) warns against overlooking or underestimating the importance of this step. A failure on the part of IT-staff and the sponsor to understand the current state of affairs can lead to discrepancy and redundancy. Evelson stresses the need for a full “BI-diagnostics” in this stage.
4. Create a data-storing strategy
The majority of BI approaches begin with the construction of a physical, isolated data farm (or data mart) since it is, at least traditionally, the quickest and easiest way. However, this creates storage issues and may potentially require further hardware additions as storage fills. New advances in cloud technology have moved these data farms off premises in a bid to save hardware and electricity costs.
5. Understanding users’ needs
The most sophisticated BI implementation is useless without an understanding of how the different employees will use it. There are three classes of users: strategic, tactical and operational users.
- Strategic users are generally high-ranking people who use BI to make few decisions, but those decisions have a lot of consequence. The decision to close a site in one location and replace it in a different country is an example of a strategic decision.
- Tactical users are likely to use their information to make decisions many times throughout the week, therefore it is likely that these employees need up to date information.
- Operational users are front-line employees such as staff at a call centre. Understanding the different uses of BI will supply employees with the relevant and recent information necessary to make an educated decision.
6. Decide whether to buy or build the analytical data model
This depends on the intentions and strategy of the BI practice. An out-of-the-box, industry-specific solution may be suitable, especially in a homogenous IT setting that has one program for CRM, and one program for ERP, and so on. Conversely, customization can be a great tool for businesses, though Evelson states that it may be simpler to start off with a standard template and customize from there.
7. Consider all Business Intelligence components
Components important to the success of a business intelligence practice are: metadata, data integration, data quality, data modeling, analytics, centralized metrics management, presentations (reports and dashboards), portals, collaboration, knowledge management and master data management. Furthermore, this requires defining and understanding the infrastructure of the BI-stack even if it is not directly relevant to the strategy itself.
8. Choose a systems integrator
BI implementations rely on guidance from an experienced partner. Evelson calculates that between $5-$7 are spent on services for every $1 that is spent on software. Further, he stresses high collaboration between end users, analysts and developers, meaning that the fine tuning of the deployment should not be outsourced. It is the difference between being given a fish and learning how to fish.
9. Start with the “low-hanging fruit”
Start with single, high-value components such as sales analytics. This is an area with plenty of solid out-of-the-box tools and templates already in existence that will not need to be customized in order to be effective. Doing this can not only add value to the company, but create a favourable perception of BI in other areas of the business, motivating further expansion.
10. Work in “baby-steps”
An end-user, analyst and developer should work together to create a proof of concept in the preliminary stages. The team should choose a few simple performance indicators and generate a few reports before growing in sophistication. The fundamentals are key.
About M.R.S. Company Ltd.
We are a full-service, relationship-centric IT consultancy firm offering our clients a compelling on-shore alternative to building and delivering best in class technology solutions. We bring experience, expertise and software development processes, refined during 35+ years of diverse engagements, to deliver complete IT solutions that significantly improve operating efficiences.