Data Enablement

Top 6 Reasons Why BI Projects Fail

Why do BI projects fail?

In today's data-driven world, the adage "knowledge is power" has never rung truer. However, harnessing that power effectively can be a daunting task, especially in the realm of analytics projects. Despite the potential insights they promise, many of these initiatives falter, often due to a myriad of challenges. Understanding the reasons behind their failures is crucial for any organization aspiring to leverage data for strategic advantage.

In this comprehensive guide, we cover the top six reasons analytics projects fail, shedding light on common pitfalls and offering actionable insights to ensure success in the pursuit of knowledge through data.

Interpreting BI and Analytical Project Failure

What does "Failure" really mean when it comes to BI and analytical projects? 

fail·ure / noun / Commonly defined as nonperformance of something due, required or expected; a subnormal quantity or quality; an insufficiency. 

Gartner estimates anywhere 70 to 80 percent of business intelligence (BI) initiatives end up failing.* But what does that really mean?

Many companies can overlook intermittent failures if the end result exceeds expectations. How often does this actually happen when it comes to BI and data strategy, though? After all, what’s deemed a success from a technical perspective still can fail miserably if it doesn’t meet the expectations of business users.

Depending upon how long the project takes to complete, there’s also the opportunity cost paid by the company while it waited for critical insights to be delivered. The longer the project takes to complete, the higher that opportunity cost. Taking this pain point into account, at what point, or at what dollar threshold, does an active analytics project fail?

Yet, at its very essence, the failure of any analytics or BI project comes down to one thing: an inability to deliver critical, up-to-date business insights in a timely manner any user can leverage to take immediate action that positively impacts the business.

* CIO Magazine. "4 Reasons Most Companies Fail at Business Intelligence." Sept. 17, 2017.

Reason 1: Underestimating Analytics' Business Value

People underestimate the value of analytics because they don’t understand its true business value.

In organizations lacking business insight, getting access to a visually appealing dashboard, instead of receiving yet another Excel spreadsheet or PowerPoint graph, may seem like progress. It is also likely that this dashboard includes too many KPIs “just in case.”

5But what exactly are these dashboards and visualizations telling you? What specifically are you going to do differently or better because of it? If you can’t answer these two key questions, you’re not asking the right questions with your analytics projects. And you’re not alone.

A BI solution has to rely on clear measurement of ROI compared to the system cost. The BI solution should also include the right KPIs to help its users answer revenue-critical questions on the spot. For the most part, people still haven’t fully figured out how to use data to make timely and actionable business decisions. Despite referencing many parameters and a lot of data points, the majority of decision-makers still rely upon gut feelings and instinct.

Often, they take this approach because they were burned before by making a purely data-based decision using information that turned out to be inaccurate, incomplete, or out of date. Because of this past betrayal, they’re reluctant to disregard their intuitions completely despite what the data may tell them.

Ask yourself:

What exactly are your dashboards and visualizations telling you?

What specifically are you going to do differently or better because of it?

Reason 2: Business Users Distrust Analytics Data

Business users don’t trust the data delivered by the analytics project.

Most enterprise analytics projects summarize data in aggregates and present the information via dashboards and visualizations. As a result, a large part of the BI market focuses on the visualization side of the equation: presenting complex information in a visually appealing and easily digestible manner.

While these tools are helpful snapshots of data they don't enable business users to easily drill down into the data, from top level all the way through to transactional details, to verify the information or understand additional context when needed.

It's nerve-wracking for a business user to make critical business decisions based on limited data they receive from a dashboard.

Most of the analytics projects that fail are the ones that focused too heavily on the visualization component. The analytics designers focus on the visual appeal of a dashboard, for example, instead of ensuring the actual business criticality of and easy access to the analytics, data, and insights delivered by it.

Some questions to keep in mind when working on BI projects are:

Where does the data come from?
What is the validity?
Do the data sets make logical sense to analyze?

The data needs to be sanitized/filtered, in order to achieve business objectives. Companies collect mass amounts of data and have numerous ways to analyze it. Focusing on the target allows this process to become simplified.

Without a fixation on the target goal, “data paralysis” can occur and the objective could be lost. Data paralysis is defined as, “over analyzing a situation to where no decisions are made, causing any forward movement to halt.”

Reason 3: Users Lack Insightful Questioning

Users don't know how to ask the right questions to deliver insights to the business.

Using the traditional, data-modeling approach to analytics, business users need to provide their business requirements and define the questions they want analytics to answer at the beginning of the project. Yet many business users are new to analytics; at the beginning of the project, they can’t articulate their business requirements, and they don’t know which questions they want to ask.

They struggle to understand what data can do for them and they simply don’t know what’s possible. Instead, they are forced to decide the questions they want to ask before they have any insightful answers from their data. The dashboards, reports, and insights they receive are often insufficient and quickly become outdated.

The business evolves, and the data evolves, but the dashboards, reports, and insights built with rigid, inflexible tools become stagnant and are difficult to modify—and impossible to update without heavy involvement from technical resources already buried by other priorities. As a result, users rely upon these once-valuable tools less and less, and the tools’ relevancy quickly dwindles.

Reason 4: Project Leaders Prioritize Technical Skills

Project leaders prioritize technical skills over business acumen.

Analytics projects powered by data warehouses and data lakes require the involvement of technologists who know how to design and implement these complicated components. As a result, many companies tend to prioritize technical skill sets over business acumen when putting together the analytics project team, but successful analytics projects need technical skill sets and business acumen to be on equal footing when setting the vision for the analytics project.

That’s because technical resources rarely comprehend the business context of what users are trying to understand or achieve with the data. Only the business side of the equation understands the compelling questions, problems, and challenges the analytics project needs to answer, and more importantly, why they need to be answered.

Many companies jump head-first into BI projects with the mantra "if you build it, they will come." Keep in mind that a good chunk of the effort in a successful BI project comes long before the actual migration occurs. Spend some time answering the question, ‘What decisions do we want to make with this data?’ You may find right out of the gate that the answer to such a question will vary greatly throughout your team. Use disagreement to your advantage; your confidence in the answers to that one question will ultimately determine your success.

Although BI and data automation tools need to solve certain problems for the organization, they also need to be user-friendly. For example, if the BI supports dashboards, these should be mobile-friendly and easy to learn in a few intuitive iterations. Most end users will not necessarily be tech-savvy. They can be clerks, sales reps, ticketing agents, call center operators, nurses, and so on, whose jobs are to focus on the customer’s needs, not the system. If the solution is challenging to use, it will only create frustration and a bounce back to the previous way of doing things.

Reason 5: IT-Business Communication Gap

There is a lack of communication between IT and the Business.

Miscommunication can often jeopardize project success. It stems from failure to comply with ambiguous requirements involving a project.

Some examples of miscommunication are…

  • The target isn’t made apparent
  • Insights from the data aren’t clearly stated
  • The users’ wants and expectations aren’t known

The blame usually falls upon the IT team because of their inability to communicate. Under 10% of IT leaders believe that they effectively communicate with non-IT colleagues. There is also a disconnect on who can have access to what data and the business does not know how to collaborate with IT, leading to data or information silos.

When you give everyone access to the data, it can make your operations more streamlined and efficient since those who know their business won’t have to wait for data scientists to interpret the data for them. The bottleneck is removed.

There can also be fault found in management for not promoting effective communication. For instance, creating dedicated communication activities or displaying the importance of IT communication. Project metrics sometimes aren’t in place, so the expectations aren’t stated. There is also an unrealistic expectation from management that there won’t be delays. Most people don’t work at full capacity every single day of the project, so a realistic timeline is important.

Reason 6: Lack of Support for Data Literacy & Culture

There is no support to improve data literacy or build a data culture.

Most BI projects are treated as "a one-time deal”. We all know that change is constant, and with time the business changes, rules change, and end users change. Companies must make BI and analytics a part of an operation that covers change management. This will help improve and enhance existing BI applications with respect to the changing requirements and functionalities. If this is not done, then in 6-18 months the adaptability of the application will decrease and spiral down.

It doesn’t matter how much data your organization collects; data must be analyzed to get actionable insights before it provides business value.

According to a Gartner Annual Chief Data Officer (CDO) Survey, poor data literacy is one of the main roadblocks to the CDO’s success and a company’s ability to grow. To combat this, that same study shows that 80% of organizations plan to implement specific initiatives to overcome their employees’ data deficiencies.

Data literacy propels forward momentum and success. It’s essential for everyone in an organization to understand why it’s important to improve data literacy. Employees should be able to use data to influence their day-to-day activities as well as big-picture decisions. If used the right way, it can help every employee achieve their objectives, perform their job better and contribute to overall company performance.

Looking at a Modern Analytics Platform

Advanced data solutions promise to overcome the reasons why analytics projects currently fail and help organizations be ready for Artificial Intelligence (AI) initiatives. But to truly be modern, the analytics platform—at a minimum—must:

  • Clearly, quickly, and consistently demonstrate business value to the people who need it most (the business users).
  • Deliver up-to-date, accurate data that’s a “single source of truth” all stakeholders can trust.
  • Liberate users from having to decide the questions they want to ask at the onset of the analytics implementation, before they’ve even received any insights.
  • Provide fresh user dashboards, reports, and insights featuring current data, whenever and wherever they’re needed.
  • Prioritize ease of deployment and implementation so much that business acumen becomes as equally coveted as a prized skillset for analytics project leaders as technical know-how.
  • Ensure the data, analytics, and insights delivered are not only presented in a visually pleasing manner, but are ones that matter to the business and help drive critical decisions made by management.
  • Not require heavy customization, intensive resources, and complicated documentation in order to properly support and maintain BI over time.

To avoid the failure and lingering pain common with legacy analytics projects, remember these seven key requirements for a "modern" analytics program. Instead, exceed the high expectations of your business users while gaining the respect and trust of the C-suite.