Avoid the Pitfalls: 3 Reasons 80% of BI Projects Fail

What stands in the way of a company fully adopting a business intelligence (BI) solution?

In a survey in his study Why BI Projects Fail – And What To Do About It  Timo Elliott referenced a study asking business technology professionals a similar question.

The findings: 51% of the respondents said integrations and compatibility issues were the biggest roadblock. 48% said ease-of-use problems. And there was even a 45% tie between not establishing a clear way to measure the ROI and problems with data quality.
Avoid the Pitfalls - Why BI Projects Fail

So why go through the effort of implementing business intelligence into your company?

Business intelligence has become an integral part of keeping a company competitive. ComputerWeekly.com reported on a Forrester survey finding that of the BI users who responded:

    • 43% have improved decision making
    • 36% improved data quality and consistency
    • 34% achieved better business transparency
    • 32% improved business performance across departmental silos

Then how can I prevent a BI implementation failure in my company?

Knowledge and planning can go a long way when it comes to business intelligence. Simply knowing your company’s needs, understanding the deployment process and having a well-defined plan can prevent many of the small bumps that could derail the whole project. Dimensional Insight created a Triangle of Success that puts most implementation issues into three categories: Process, Technology and Management.

Process Failure

Read reviews. Find an all-purpose BI solution. Implement into company workflow. – Simple as pie, right? Wrong.

A process failure involves the reasoning and the strategy behind your entire business intelligence project.

Finding a Reason

Before you hop online and start looking up vendors, you must find out why your company needs a BI solution. Without a well-defined goal, the BI project will get out of hand quickly.

To identify the reason to start a business intelligence project, you have to get the entire company involved. Talk to all users who will be affected by the solution and ask questions about what works and doesn’t work with the current distribution of information, data quality, decision making process and report creation.

Pitfall Prevention: While talking to users in each department, make sure the whole company agrees on the definitions of key terms and business metrics to eliminate confusion.

While wading through the responses, make a list of the issues that need to be addressed starting with the most critical to the company. That will become the implementation objective. However, don’t get rid of the list once you determine the goal– you’ll need it for future BI projects.

Next, it’s smart to begin looking at business intelligence tools and features. Start creating a list of use cases so you know exactly what the company needs from a tool when selecting a vendor and solution.

Planning a Strategy

The strategy of implementing a tool is the backbone for the entire BI project. In their 9 Fatal Flaws to deploying a BI solution, Teradata Magazine labeled a poorly developed strategy as the “biggest flaw” a company could have. The more thorough your strategy is, the more chances you’ll have at a successful company-wide adoption of the solution.

When beginning to lay out a business intelligence plan, a company should look into:

      • The best deployment method for the tool.
      • How much and what type of training will be needed for users.
      • How to measure the return of investment (ROI).

By talking to users within the company and identifying goals, you’ll get a feel for some of these points, such as whether your company will have a better adoption rate with a phased approach to implementation or an all-at-once one.

Pitfall Prevention: For fewer surprises, create a strategy that includes every step of the process. This not only allows you to plan ahead, it also gives a company more of an understanding of how big of a budget the project needs, how long it will take to adopt the solution and how to gauge the ROI at the end of the process.

Technology Failure

A technology failure is almost self explanatory. It involves any technology problems a company may run into during implementation, from selecting a solution ill-suited for the company to collecting poor quality data.

Selecting a Tool

Finding the right fit with a BI tool isn’t easy. In fact, Teradata Magazine mentions it’s a “tough juggling act” between meetingthe current objective and planning for the future.

When looking at the features and functions of business intelligence tools, consider what users in all levels of the company will need. A CEO may need quick snapshots of decision-making data, while managers and supervisors may need to drill down into each data point. Understanding who needs what type of information and how it should be presented is only half of finding the right solution.

The other half is finding a solution that will foster growth within your company. When selecting a BI solution, your main focus should be the well-defined objective determined at the beginning. But remember, you made a whole list of issues that need to be addressed. However, you shouldn’t look for an all-purpose solution right from the start. You should be aware of how well the solution can to handle more sophisticated needs down the road.

Pitfall Prevention: Choose a business intelligence tool that will evolve as your company grows.

Data Integrity

Data integrity can be compromised either by collecting poor quality data for the desired insights , or by a lack of understanding of what the data is trying to say.

There is a lot of data out there that can offer actionable insights.

But collecting and analyzing all the data available to find a specific insight is not only tedious, it can also cause problems with data quality. Before jumping into a pool of thousands of data points, find out what sources you’ll need to collect data to find the desired insight. This not only keeps you more focused, it also prevents you from using duplicated, incomplete or outdated data (also known as “dirty data”) as well.

Data integrity can also be compromised by not understanding what the data is saying.

StatSlice touched on this in its survey that asked business intelligence professionals about the top mistakes organizations were making in BI and analytics. An example of a lack of data understanding can be as simple as misunderstanding the rule sets dashboards use to present data. If this occurs, “the same data may mean different things to different people.”

This problem can be prevented by making sure all users understand the technology’s rule sets and metrics of collecting and presenting data.

Management Failure

Having proper leadership to manage the BI project is one of the most important keys to preventing failure.

Developing a Team

Many times, a business intelligence project is left solely in the hands of the IT department or business executives within the company. This can cause problems with the solution due to the lack of understanding of what’s needed on both sides.

According to Domo’s How Business and IT Professionals View Business Intelligence infographic, the two types of professionals have different priorities when it comes to what is needed in a BI tool.

Business executives look for:

      • Ease of use
      • Speed and agility
      • Clear and concise information
      • Sustainability

IT professionals are concerned about:

      • Solid security
      • Access controls on data
      • Compliance with regulations
      • Ease of integration

The different perspectives also change how each view the success of the business intelligence project. In the same infographic, 24% of business executives believed the BI deployment/performance was very successful, versus 32% of IT professionals.

It’s best to create a team of experts from both the IT and business departments to lead the BI initiative. This can prevent potential limitations in the adoption rate of the tool.

You might find the BI project getting derailed if you don’t gain support from executive management. Without a full understanding of how the technology will improve the bottom line for the entire company, preconceived ideas about outcomes and future roles could stop a BI initiative before it really gets started.

A respondent to StatSlice’s survey offered an example of how an influential department manager shut down a BI project because he believed the solution would undermine him and his team’s function of manually producing reports.

Pitfall Prevention: Educate all managers who will be impacted by the new BI technology about how it’ll improve different positions. This can prevent political issues within the office during implementation.

Project Management

Once a team is in place, it can manage the objective of the BI project and determine whether it has too large of a scope.

If an objective or issue is too large to tackle, the team should consider breaking it down into smaller goals. For example, a company may want to increase its overall revenue. This can be broken into more digestible projects such as converting one-time customers into loyal ones, increasing customer loyalty purchases and identifying new niche audiences the company can target.

How Kraft Succeeded

As the world’s second largest food company, in 2010 Kraft Foods found it needed to update its business intelligence system to drill down into data better and discover trends or issues more efficiently.

As a company, Kraft decided to retire the legacy applications it was working with and try using SAP BusinessObjects BI 4.0. Heading the project was Mike Walsh, Associate Director of Business Intelligence at Kraft, and Kelli Such, Kraft’s Global BI Director.

Up until then, the focus had been “on implementing SAP transactional systems at various Kraft plants and businesses, getting the configurations right and ensuring users were properly trained on those systems.”

This meant the company’s priorities were set on building off the data they already had in SAP NetWeaver BW and using traditional BI reporting tools to provide visuals.

Sustaining over 70 brands with $100 million in revenue, more than 40 brands that are over 100 years old, and a handful generating over $1 billion in revenue, Walsh and Such knew that Kraft’s businesses needed to see a quick ROI to jump on board. Although they believed the transactional systems were bound to have positive results, the two BI project leaders knew the best way to quickly show the benefits would be to deploy user-friendly BI applications across Kraft.

In the SAP case study, Walsh said, “We needed to produce the types of reports that helped people make business decisions. These kind of reports make it obvious to people why the decision was made to go with SAP software.”

To get to deployment of the SAP solution faster, Such and Walsh came up with a four-point plan:

  1. Before moving forward, gain full support of the IT organization, including financial support.
  2. Develop BI capabilities in advance of business demands and requests, and then showcase commercialized product.
  3. Create a dedicated group to develop BI solutions, rather than pass demo versions to the businesses.
  4. Sustain a close relationship with SAP to streamline deployment and work out bugs.

Going through these four steps, Kraft was able to create a successful BI model for whenever a new solution is introduced. A BI Technical Service team was formed under the supervision of Walsh to guide development and deployment of the tools before the business units see them.

Walsh’s team focused on the transports, notes, connectivity, security, internal controls and cost model so when the solution is given to the business units, implementation can be done quickly and efficiently. “Empowering the businesses to do this drives the adoption rate and speeds up the turnaround time,” he says in the case study.

Aside from the speedy deployment, the Technical Service team also helped the businesses with adoption. This was done by showing the users the tool before beginning the implementation project to get everyone on board from the start.

By dedicating an immense amount of time, effort, and resources to plan and foster its BI project, Kraft had tremendous results. Walsh stated, “Users went from reports that took weeks to create, to reports that took minutes to create. This gives them a brand-new user experience that they probably didn’t even think was possible.”

Selecting The Right BI Vendor:
The Ultimate Guide

Choosing a BI vendor is all about finding the right fit. Our exclusive report will walk you through the process and help you select the perfect solution.Download Now

Comments

  1. “How long it will take to adopt the solution and how to gauge the ROI at the end of the process.”

    Measuring the ROI of your BI project is imperative, and yet many businesses aren’t able to proactively track ROI as they never set up benchmarks when they launched the project. Establishing goals at the launch creates a baseline for companies to compare their BI performance against – which in turn can provide a better sense of the return on investment.