Business Intelligence Fundamentals – Best Practices that Bear Repeating
Rather than waste time debating the right or wrong way to approach a Business Intelligence program – just keep the start small, focused and simple.
You may have heard this before, but it is worth repeating. Starting small and scaling up after capability is proven is a fundamental consideration when designing your Business Intelligence Program, as are:
- Targeting well defined business and technology outcomes and capability requirements
- Understanding the source and quality of your data
- Having an informed view, not just a vendor sales view, of Business Intelligence tools
- Ensuring security measures and governance are in place from day one
- Assembling a Program team with the right skills and experience
The above guiding principles are the basics but they’re anything but basic. The most successful Business Intelligence programs we have delivered, many of which were start up investments in large organisations, take the time to consider these elements carefully at the outset, to agree on priorities, and sharply define what success looks like – and then be prepared to change.
Many, if not most, business intelligence projects will undergo either superficial or fundamental changes over the course of the program. The perfect definition of the business problem, the perfect source of data, the perfect understanding of how to accumulate and aggregate data, the perfect Business Intelligence tool, are difficult to nail down in the real world.
Even when the business investor thinks they have clearly defined the targeted business outcomes, once they see the initial Business Intelligence outputs there may be modifications to data sources and Business Intelligence outputs. Even if the Business Intelligence Program team thinks it knows how to generate the required information outputs, once they start using the Business Intelligence tool and once it is clear what needs to be done to and with the data, they may need to modify their original designs. Even when people think the data sources are well-defined and the data has been cleansed, often the reality is that they may end up needing to use alternative data sources than those originally planned, and more data cleansing is then required.
The beginning stages of Business Intelligence Programs always have iterations. The aim should be to get changes dealt with early in order to minimise rework. This is why we recommend starting small and then scaling.
A small list of well-defined targeted business outcomes and, depending on the program’s intent, a starting team of five to ten people can generate targeted deliverables quickly with acceptable risk and investment. A small team can bring to life the importance of the above guiding principles and what is needed to deal with them.
The most successful Business Intelligence Program teams include:
- A business subject matter expert who knows what they want
- A Business Intelligence expert who knows how the tool works
- A data expert who understands the underlying data and its specific challenges
- A fit-for-purpose project manager who understands Business Intelligence delivery
Imagine this core team starting with one or two target outcomes, with a clear understanding of the data sources to be used. Such a simplified start to your BI Program allows the team to uncover many of the challenges that need to be dealt with before the program grows. Then, once Business Intelligence delivery capability is proven, scaling can be assessed and a considered investment program can be designed.
In terms of extending the team it is worth noting that, while most organisations set up Steering Committees, Stakeholder Working Groups and Program Management controls, Business Intelligence Programs call for a couple of additional stakeholder feedback opportunities. Adding Data Governance Forums and Business Initiative Prioritisation Forums can help to drive real and continued business value out of the Business Intelligence investment.
In the second part of our Business Intelligence Series we take a closer look at dealing with data.