According to research from Gartner, between 70% and 80% of corporate business intelligence projects fail. Why is that? How can you prevent failure and be one of the 20% to 30% that succeed? We want to help you! In this post we’ll outline 4 common challenges of BI implementation projects and how you can overcome them.
Driving User Adoption
Many organizations operate under the impression that if you build it, they will come. This is a fallacy. Analytics adoption takes real work as people won’t want to change the way they do things unless their current methods are tedious, time-consuming, or just don’t work. Even when inefficiencies are recognized, promoting change can be difficult, especially when individuals are averse to risk.
For example, the University of Chicago knew it would encounter resistance to adoption during a recent business intelligence (BI) project, impeding two goals: increasing stakeholder input in design and encouraging stakeholders to become advocates for future expansion. To encourage adoption, they launched an internal training and marketing campaign and built a series of posters, emails, and interactive videos with the goal of getting employees excited about use. When the solution was rolled out, the end user base crept up to over 400 people even though the initial intention was 35 users.
How to overcome it: to avoid resistance to adoption, it’s important to involve your stakeholders THROUGHOUT the implementation, not just at the end. Get their input on design, user experience, and benefits early on; then show them how you’ve implemented their suggestions in a dashboard that will make their jobs easier. By the end of the project, early adopters should feel a sense of ownership and become your strongest advocates. Whether saving time, gaining better visibility, or supporting job functions, employees need to feel empowered and this includes designing their own interactions with the technology they are required to use on a daily basis.
Building What’s Right, Not What’s Easy
Sometimes, developers are so inundated with business user requests that they end up creating all-purpose solutions instead of the right solution for the right audience. But this kind of behavior only leads to more frustrated business users and project reworks. What is the old adage? “We never have time to do it right, but we always have time to do it twice.” Resist the urge! Commit to solving the real business problem, instead of crossing off the project from your to-do list. Even if multiple users (or user groups) need access to overlapping data, their interactions with it will probably be different. Thus, you need to isolate each user group and identify their business challenges so you can design solutions that correctly match their needs, and deliver it where they need it.
How to overcome it: start small and focus on just a few KPIs for each user group. As you deliver intelligent tools and users become proficient with them, they will begin to ask new questions and adoption will increase over time. In the beginning, make sure you’re solving problems they face now and make it easy to enhance functionality in stages. Dashboards are meant to change with your business over time, enabling you to remain informed as your business needs change.
Justifying Your Investment in BI
Another challenge organizations face is being able to justify the cost of developing new dashboards. Organizations with mature BI infrastructures will often be reluctant to change, add to current investments, or shift perspective on existing business processes. Traditional BI deployments were generally based on the development of large IT infrastructures, which delivered information inefficiently. The transition from traditional BI towards more agile dashboard solutions will require a new approach to BI.
New investments might create a challenge in identifying ROI and whether it outweighs the risks and/or expenses. After all, explaining the value proposition of dashboard adoption may be difficult if past BI investments have come up short. Although this may be the case for some organizations, others, such as Pathway Chicken [name changed for privacy reasons], a $5 billion poultry product producer in North America, illustrate how to measure quantitative ROI of BI investments. To support its just-in-time production strategy, Pathway Chicken developed a set of four dashboards that displayed near real-time information on temperature, pick-up, weight, pack, etc., which allowed them to determine if they were on target on the plant floor. Because of the bold graphics and rich colors they selected, they were able to achieve a high level of adoption and have estimated savings of more than $2.4 million annually by being able to better respond to information affecting their operations quickly.
How to overcome it: just as Pathway Chicken did, use your new dashboards to demonstrate real-time improvements in workflows and time savings. If you take the time to create comprehensive and effective dashboards, the data and visualizations should speak for themselves. Use vibrant colors and charts so all stakeholders can clearly see the improvements made by creating better informed users.
Making BI Accessible to All Skill Sets
Dashboards have the potential to grant access better information to all employees, so they can make more informed decisions and increase productivity. The problem, however, is that designing dashboards for a wide range of business intelligence users, with varying comfort levels with technology, can be a challenge. After all, some users are very familiar with interacting with BI while others are new to the concept. Organizations looking at deploying dashboards are tasked with developing solutions that can be used by employees with diverse skill sets.
How to overcome it: effective dashboard delivery requires accurate and timely information delivered in a way that is digestible by users of all comfort levels with technology. And this is more challenging than it sounds. It may mean providing more than one dashboard with different levels of interactivity and data access. Or it might mean looking at the same dashboard in different ways as different people may require the same information for different purposes or with disparate latency needs.