Eight reasons why business intelligence implementation can fail

BI tools can be very beneficial to an organization by delivering helpful insights, but common mistakes frequently derail their effectiveness and render them useless.

When it comes to business Intelligence implementations, the common driver for adoption is the promise of delivering insight to the organization through data-based decision making. This allows many groups to take the guesswork out of any risky choices. If business intelligence can offer this acumen to companies, why are some abandoning it completely to revert back to traditional reports and with complex and non-user friendly reporting engines?

There are a number of contributing factors that negatively affect the outcome of BI initiatives. While they can differ from industry to industry, here are the common business intelligence pitfalls that should be avoided:

  1. BI should never be considered a one-time project. Unlike traditional IT projects, such as server upgrades and migrations, business intelligence implementation does not have a hard end date. As business users begin to utilize dashboards and key performance indicators (KPIs) and begin to apply different levels of analytics, the BI initiative continues to grow and adapt past its rollout date. Data begins to grow and end users adopt the BI tools available to them. Initially, most users focus only on working with KPIs and dashboards, but they can move past this to exploit predictive models and machine learning capabilities that are available in some platforms.

  2. One dashboard does not fit all users. Despite the fact that many organizations will have common business challenges, they will most likely have different BI needs. This applies to the different kinds of users within the same departments or organizations. For example, nurses and providers work together on patient care, but will likely look at different visualizations and data points. In addition, one hospital's financial system or overall strategic objectives will differ from another, and that requires unique KPIs and dashboards.

  3. Flashy charts are not always able to deliver the right insights. Data may be one of those critical areas that, if not executed correctly, may not be able to relay the appropriate insights. This results from a lack of understanding about how to map what data to which charts. Some BI vendors focus too much on showcasing different charts, colors and effects, but miss the most important aspect: the right chart for the right data.
Most users focus only on working with KPIs and dashboards, but they can move past this to exploit predictive models and machine learning capabilities.
  1. High licensing costs can mean narrower adoption. BI should no longer target only executives and managers; it can provide valuable insights to business users at all levels. But high licensing costs have historically forced IT to purchase only a limited number of licenses in order to keep costs down and focus on delivering those tools to upper management. Fortunately, as more BI platforms have been offered as a service, pricing has become subscription-based, making BI tools more affordable.

  2. BI is not an overrated tool with no real value. Users that tend not to buy-in from end users is the result of limited understanding of what BI can do to help an organization transform itself. Business intelligence is the discipline of evaluating data collected and used by organizations and analyzing it to gain insights into what is affecting the business, as well as what areas must be improved upon within the business.

  3. BI is more than just historical data. A common use of BI is to look at historical data in order to evaluate previous events based on available data. That is not enough to truly showcase the full capabilities of BI stacks. Business leaders, managers, supervisors and standard users can take advantage of what BI has to offer in terms of what is happening in real time as well as in the future.

    Organizations can use their historical data by applying different machine learning features to help predict upcoming trends and be better prepared for what will likely occur in the future.

  4. Different data sources should never stand in the way of BI. Over the years, many organizations have gone from having data stored in different data formats on their servers, to having data stored in different data centers. With the rising popularity of software-as-a-service-based products, many organizations have realized that sizable portions of their data are in the cloud, and the traditional BI platform may not be able to connect to those services in order to make meaningful use of it. Since there is more value in connecting all organizational data together to have better visibility of what is occurring within an organization, it is imperative to look for BI tools that support multiple sources and ensure that the connection between these different information silos is available. An example of different sources could be a payroll service hosted online, a sales CRM system, an internal ERP tool and even social media platforms.

  5. BI is only meant for executive roles. The BI sales pitch tends to focus on showcasing the platform and its flashy visualizations to executives, but in reality, the platform should be offered and explained to members of an organization at all levels. Managers, supervisors, team leads and other employees can gain tremendous value if they leverage many of the capabilities that BI has to offer. While some executives need to have a wide view of the current organization's sales trends and finances, sales reps or even folks in the production line should have the option to review reports and dashboards that reflect their overall performance and ensure they are on track.

As organizations continue to try to streamline their operations, find ways to grow their businesses and be more efficient, using data to gain insight becomes a critical task. Whether it is through pulling their existing investment into their BI solutions or adopting a new one, IT executives must ensure that any new attempt at BI must avoid the common challenges and pitfalls that can cause the project to fail.

Next Steps

The rise of BI in healthcare

Power BI tools provide accessible data

Mainstream BI implementation remains top priority

Dig Deeper on Collaborative analytics platforms and strategy