6 Ways BI Projects Go Wrong and How to Avoid Those Pitfalls

Everyone who has been involved in a BI implementation knows that there are many factors that can prevent success.  A myriad of problems can arise that seemingly prevent the project from completion and providing value.  Here we cover six of the most common project issues that arise and how to effectively overcome these obstacles.

Before you tackle any BI project, you need to think about your overall data and analytics strategy. Do you have one? Consider these questions before embarking on your project: How does this project fit into your overall analytics goals?  What about your company’s overall corporate strategy?  What is the business reason behind the need for analytics?

Solution: We recommend developing an Analytics Plan that will help align each of your analytics activities with your overall corporate goals and then determine the technical feasibility.  This will help you assess the current state of your analytics practices, prioritize efforts, and build a long-term plan so you can start turning your data into valuable insights.

How often have you heard the story of the executive team that has purchased a software package and expect it to solve all their issues?  From the hype told by the software salesman, your shiny new BI software package will do everything you need it to do, implementation will be easy, and user adoption will soar.   It becomes easy to buy into this narrative and expect all of your data issues to go away with the right software.  While there are many great technologies that help achieve your analytics goals, there is rarely a “cure-all” solution.

Solution:

>>  First, identify the business reasons for wanting the technology.  Too many people focus on the “cool” features that they likely won’t use, when they need to ensure the technology answers their business problems.

>>  Evaluate each potential software solution thoroughly.  Insist that the vendor complete an involved proof of concept based on a real issue your company is currently struggling to solve.

>>  Request that the software vendor provide a detailed and thorough product demonstration.  Make sure the demonstration is well attended by employees across all departments, and that you ask pointed, difficult questions.  Do not stop inquiring about the software features until you are satisfied you have reached a truthful answer.

>>  Evaluate all possible vendors completely (even if you have a favorite), and request referrals from other clients.

 

After spending much time and effort to scope out and plan an implementation, you discover that the persons you thought were key to the project success have little ability to help with influencing adoption, budgetary issues, or determining application requirements.  This often happens when projects are developed and not shared widely within the organization.

Solution:

>>  Complete the project in sprints (or increments), and demonstrate the results of every sprint to the widest audience possible.  The more persons who see the application along the way, the sooner you can determine if there are other key decision makers who should be part of the project.

>>  Strive to include IT AND the business user community in your project.  Make sure all departments are represented to review and give feedback on your sprints.  This also helps identify project champions who you’ll need for a smooth implementation.

>>  Show real value in each sprint.  Once people see how the project affects them, they are encouraged and inspired to get involved and give constructive feedback.

 

Many projects start with plans to “build a small simple home,” but somehow over time morph into an attempt to build the Taj Mahal.  It is very tempting to accept each different stakeholder’s small addition or “improvement”, but this puts you at risk of enlarging the scope to such a point that nothing of value is delivered, and timelines are impacted.

Solution:

>>  Break each project down into the smallest sprints possible that deliver business value.  Stick to the original plan for the sprint you are working on.  If a change is requested, complete the original sprint and submit a change request for the additional functionality.

>>  Limit the moving parts to each sprint.  Always make the deliverable as simple and straightforward as possible.  Resist the urge to “boil the ocean” and accomplish everything in one sprint.

>>  Set realistic timelines. Resist another urge to “get it done quicker”. Quality over quantity will benefit your project in the long-term.

 

Talented persons with good skill sets are always in high demand, and it is tempting to “loan” them to other projects that have an immediate need.  You may be promised that the person will only temporarily be engaged in another effort, but this can turn into a long-term absence which jeopardizes the success of your BI implementation.

Solution:

>>  Have a plan in place to keep employees with the right skill sets on the project until its completion.  Resist the urge to loan people to other projects or to disengage them before all development, testing, and roll-out has been completed.

>>  Spend time up front finding the right team members.  Don’t skimp on the time required to assemble the team best suited for the job.

>>  Have a backup plan and budget if it becomes necessary to bring in outside consulting or other internal resources.

 

If no one uses the BI application you have built, then all the effort made will be for naught.  All the money and time spent won’t benefit the company or its bottom line.

Solution:

>>  Include the business in the decision of what software is purchased.  Strive to get buy-in so that the interests of end-users, business stakeholders, and IT are aligned.

>>  Again, include members of the end-user business community in project development.  Don’t develop the application in a vacuum and then expect it to be adopted.

>>  Use the Agile project management approach to deliver iterative content in small doses.Encourage thorough testing and feedback at each iteration to ensure the application solves their problems.

>>  Monitor usage.  If usage doesn’t remain strong, try to determine why.  Make changes and address issues as needed to avoid negative impressions that affect long-term adoption.

>>  Bottom line – focus on delivering value to your customers.  An application must be viewed as useful from the beginning.

The six points in this article are a good start to putting your project on a path to success.  When building your next BI application, be sure to include the correct decision makers, be realistic about software capabilities, avoid scope creep and changing team members, and pay close attention to adoption rate.  By doing these six things, you can put the odds in your favor of having a successful delivery.

Analytics and big data. It's what we do.

 

Contact Us

National Office Telephone | Mon-Fri 8:30am-5:30pm CT