BI Best Practices 101
Many Business Intelligence (BI) efforts fail to reach their potential and fail to deliver on their intended result. It matters not how beautiful the visualizations are, or how much effort was required to extract the data, or even how accurate the numbers represented are. In order to be successful, the key decision-makers need to focus first on ‘why’ the business is embarking on this project, and then on the business processes and behaviors of the users of the BI solution.
Every successful project must start with a definition of what success looks like. Yet, according to Gartner, almost 70-80% of all BI projects fail. Is this a software issue? Unlikely. It’s less to do with the product, and more to with the process. Processes are where the issues fundamentally begin. Management has a vision, but sometimes they neglect to communicate that vision to IT and those executing the BI project. This lack of communication between IT and the business means the project has the potential to fail to deliver. The business will explain to IT ‘what’ it is they want to accomplish, however they won’t explain ‘why’. There is no discussion outlining specific goals the business wishes to achieve. This disconnect causes IT to form assumptions that the business knows exactly how to proceed with the project, which is referred to as ‘a suggested solution, masquerading as a business need’.
Business personnel and those administering the solution, generally will not deliberate about project specifics and what success will look like. It’s vital to a project’s prosperity, that success be well-defined prior to implementation.
Many businesses execute BI projects, but don’t have a plan to change behavior. In conjunction with BI, it’s imperative that processes – both internal and external – are adjusted accordingly. If data is purely visualized, yet behavior remains constant, then it’s nothing but pretty information; Business Art. BI projects are often undertaken despite a lack of consideration regarding how human behavior will be affected, or should be affected. Finding what the desired change in behavior is, and communicating that to all parties involved in building the project, is a key deployment success criteria.
BI drives change. It is the catalyst to giving information and insight into what behaviors should be changing. If there is no change in behavior based on the data made available, the probability of failure will be insurmountable. What is the benefit to a user, if they are unable to affect the data? More often than not, users are presented with data that they are incapable of altering, which begs the question – How can users influence behavior, if their behavior doesn’t directly affect the data?
There is a discipline between BI and human behavior. The BI disciplines are merged with human behavior, however the vast majority of businesses will keep them separate and independent of one another. To ensure success, the two must unite. The BI solution should affect the user’s behavior, in a manner that is not manipulative, but rather informative.
Quality vs. Quantity
Measuring the quantity of something, is not the same as measuring the quality of something. Moreover, it’s irrational to assume that quality would remain constant, if it’s not measured alongside quantity. If one increases, typically the other one suffers. It’s crucial that the qualitative aspect of a number is measured. Key decision-makers should determine a precedent by which to evaluate quality. By setting criteria and standards that help define quality, it becomes simpler to understand the necessary steps required for an activity to be considered of value.
Much like your buyer funnel, BI is about more than just accuracy and volume. Low quality metrics can waste a user’s time, whereas high quality metrics and data can facilitate better decision-making and behavior. For example, quarterly sales is nice to see, but if you have a 3 month sales cycle, it may be of more benefit to understand the number of leads in the funnel beforehand.
Empower, Don’t Monitor
A large proportion of BI systems are used for monitoring purposes, and not for empowerment – or for the betterment of users. A typical BI system allows management to monitor workers, more so than provide the worker the opportunity to improve their performance. What this will result in, is little, to no benefit to users, offering more hindrance above all else. A prominent cause of low adoption rates, is largely due to BI’s usage as a monitoring tool. Sufficient adoption is made possible if there are clear upsides that users can achieve and strive for. Without this, behavior will not change, certainly not for the better.
The best predictor of future, is history. This premise can be applied to BI to help recognize, that if you can see patterns and your numbers can predict the future, you will be able to better advise users on how to realize success. It’s critical to use BI to foster an environment where users will go above and beyond what is expected. Businesses need to build systems that will motivate their users, and guide them along a path to success. If the BI function is designed to encourage better achievement as a result of a well-defined path, then the organization will be superior. A BI solution that fosters personal and professional growth will help to ensure that users are nurtured and empowered.
Furthermore, it’s recommended that BI solutions should try to transition the business from one that is subjective, to one that is objective. With the objectivity that BI can bring, that must lead to a change in behavior.
If you approach a BI project with this mentality, you can expect to have a more rewarding implementation. Being able to visualize your data is great, but having a clear notion of success and having the proper attitude, will streamline the entire process.
About the Author
Jordan Zenko is the Community & Content Manager at Dundas Data Visualization. As Dundas’ resident (and self-proclaimed) story-teller, he authors in-depth content that educates developers, analysts, and business users on the benefits of business intelligence.Follow on Linkedin