Business Intelligence Challenges
Though BI has helped countless companies make smarter, data-driven decisions, it has yet to gain universal adoption.
We’ll explore the challenges businesses face while embracing, implementing, and getting the most out of a Business Intelligence solution.
Finding the Right Solution
Too often companies skip the all important step of conducting a thorough internal needs analysis. This is an essential part of the process in finding the right solution for your organization.
Up to 70% of BI implementations end up failing to meet all the business goals, according to Gartner.
But when done correctly, BI can generate an ROI of $10.66 for every dollar spent.
Before reaching out to potential vendors or even beginning your research, you’ll want to develop a set of business requirements and goals.
Consult with key stakeholders, including:
Identify the key problems they’d like to have solved. Have them develop a set of use-cases, the more specific the better.
Once formulated, categorize your list of requirements into three buckets:
- Must Have
- Want to Have
- Nice to Have
Let these guide your research process. Present potential vendors with a list of use-cases to find out how and if their solution could solve them.
These use-cases will help you better evaluate potential vendors.
The more work that’s put in upfront, the greater your chance of a successful BI implementation.
Justifying the Investment
Companies turn to business intelligence to improve operational efficiency and save money, but it has to be able to generate a true ROI.
Business leaders looking to justify the cost of an analytics system have to answer several questions:
- What are the evidence-based benefits of business intelligence and can it be quantitatively measured?
- How can upper management be convinced that it is a worthwhile investment?
- When will the company begin to see ROI?
These questions can be challenging because BI’s true value cannot be completely measured until it’s implemented.
“You can’t know the ROI, or pinpoint where it is, even though it’s there,” says Tim Biskup, Director, CRM, at Progressive Business Publications. “So it’s hard to sell at the outset.”
Generating an ROI
For example, data may reveal that warehouse shipments go out more efficiently when two workers take their breaks at the exact same time. Data will also reveal how much money can be saved by such a simple fix. But oftentimes, companies will be unaware such simple fixes exist until they implement BI. And that’s where data champions come in.
“We don’t always know up front where the benefits will be,” says Biskup. “Companies need ‘data champions’ that can explain and really sell the upside of what BI can do.”
To clarify the specific value of BI for a company, such data champions need the technical knowledge to understand analytics but also the ability to clearly communicate the benefits to non-IT users.
Key responsibilities of the data champion include:
- Building and fostering relationships between IT, executives, business users, and outside vendors
- Helping the company develop a strategic vision for the future
- Pinpointing the where and when of ROI
- Analyze a company from the inside-out and pinpoint where BI can help
Lack of Expertise
There’s no question that in order to get the most out of a BI solution, you need the right personnel.
One survey stated that 60% of companies think their employees need to develop new data-related skills to turn information into actionable insights.
Across industries, executives are striving to make their company culture more data-driven. This means empowering everyday business users with new skills.
“Companies are doing a comprehensive job collecting massive amounts of data with today’s modern data infrastructures, but the total value of this data remains elusive, says Looker CEO Frank Bien.
The Harvard Business Review has identified several traits which must define individuals working with successful data-driven organizations:
- Able to scientifically experiment to solve business problems
- Competent at mathematical analysis
- Data “literacy” skills; able to see value in data where it exists
Businesses looking to maximize their BI potential should create talent management strategies for relevant personnel to develop the skills needed for success.
Working With & Without IT
Since the modern origins of BI, the industry has suffered from a lack of technical experts.
And despite the increasing availability of simpler tools and greater educational opportunities for advanced experts in the field, masters of the trade are still in great demand. Many projects still require an expert touch.
Some vendors are responding to this challenge by changing the nature of the partnership between IT and the end user.
“There is a new emphasis on simplifying access to personal data, simplified authoring, rich visualization, and in memory technology for big data sets,” Dave Marmer, Senior VP of Information & Analytics at IBM says.
Two different types of technologies that have evolved in response are business user-led analytics that are independent from IT, and IT-managed self-service analytics.
“We see customers using both of these offerings depending on persona, use case and preferred access, and driving a partnership between business users and IT to address the spectrum of analytic needs,” says Marmer.
The insights companies gain from data are only as good as the quality of the data analyzed.
It’s simple: bad data in equals bad data out.
Bad data lingers for a few reasons:
- Multiple data sources which need to be integrated
- Enormous growth in the raw volume of total data
- Companies not maintaining proper data hygiene practices
And because companies use insights from BI to make important business decisions, bad data can cost millions of dollars.
Source systems are the main cause of data quality issues.
Companies can draw their information from any number of different data sources. Each source may format their information in a distinctive way. For example, one source might store phone numbers with hyphens, and the other without.
The most common example is personal information inputted incorrectly into the system. If the system is unable to check and correct the error, then faulty data is feeding the insights that your company is paying for.
Executing business intelligence projects often requires the creation of data categories that didn’t exist when the information came from its source.
“70% of the risk and effort in the DW/BI project comes from this step,” says data warehousing analyst Ralph Kimball in his book The Data Warehouse Lifecycle Toolkit.
BI architects are charged with resolving data quality issues through data exploration and deeper analysis of data sources. Managers should discover where these problems come from and ensure that all users are aware of their sources.