5 Major Data Analytics Missteps and How to Avoid Them
If your organization is like many others, when you look back on past data analytics projects, you probably deem them unsuccessful. Sometimes the cause was expected: delays, cost overruns, and so on. But the most common reason is more fundamental: In many cases, an organization selected a data analytics tool that ultimately failed to meet users’ needs.
That is an issue that the increasingly popular self-service analytics piece aims to address. Rather than being limited to the information you are given, self-service tools allow you to create the analytics, reports, and dashboards you need.
It’s important to note that the success of self-service business intelligence (BI) hinges on more than user adoption. I’ve noticed five major mistakes organizations often make when implementing self-service analytics. Read on to explore the top BI mistakes and learn how to avoid them.
Mistake #1: One-Size Data Does Not Fit All
A new saying is making its way around offices worldwide: “If it can’t be measured, you won’t get budget for it.” This new mentality means that people who may not have given data a second glance only a year ago are now expected to use it to not only measure results, but also to make increasingly data-driven decisions. And we are not just talking about the data analysts in an organization. People in marketing and sales, in the C-suite, and even on the factory floor now need to review, create, and analyze reports.
Herein lies the issue: All these users are not the same. They have varying skills and needs when it comes to data. As an operations manager, I need access to a lot of data, as well as the ability to play around with different data sets to figure out why certain campaigns or people are performing better than others. A sales rep, on the other hand, doesn’t need nearly as much information.
Despite the variety of users, many organizations take a one-size-fits-all approach to data analytics. They purchase a tool and provide it at the same level to all users. However, the tool may be too complicated for users who just want to review a dashboard with the highest-level KPIs, or it may be too simple for users who prefer to interact with and analyze data to uncover new insights on their own.
In order to see success with self-service analytics, you need to provide the right tools to the right people in the right ways. The better you understand the varying needs of your end users, the better you can serve their needs with tailored self-service capabilities.
Mistake #2: Incompatible Tools
Even organizations that set up different users with different tools and access levels often run into a whole different issue: adopting point solutions for each end user type that don’t work together and are difficult to maintain.
Sometimes this issue arises from data analysts going around IT to purchase a data discovery tool (I for one have been tempted to do this in the past). Other times it comes from IT looking for a new modern tool to use in addition to traditional BI.
To ensure all your BI tools are compatible with each other, select a comprehensive self-service product suite that provides a range of capabilities that can be tailored to end users roles and skills. As I mentioned above, you need to match capabilities to your users. Your information consumers want reporting with little to no interactivity while your more data-savvy users demand drill-down analysis, and your analysts crave visual data discovery that helps them figure out the answers to new questions as they come about. You need a suite of tools that addresses all of these capabilities, yet still works together to create an agile analytics cycle. This will ensure you get a clear picture of all the data across your organization.
Mistake #3: Inadequate Data Governance
Self-service BI does not mean a data free-for-all. Data governance is key to balancing the business’s need for data access with IT’s need for appropriate data security. The key is finding the right balance. Some organizations keep all data under lock and key. This leads to frustrations at many levels, especially when users want to combine data sets to find new insights, or even just create some new charts from a single set of data.
Other organizations have set up their data analytics in a way that completely ignores the need to control their data. Users can pull data from their cloud-based apps, Excel, and other sources. However, with all these various data sets floating around, they no longer have a single version of the truth.
Fostering an environment that encourages self-service while also governing the access to data in a centralized manner is achievable. As you configure your BI tools, take care to establish the necessary security controls and auditing measures that ensure users are given the right data access, and also provide IT with the transparency they need to see who accesses what data.
Mistake #4: Unclear Roles and Responsibilities
Data analytics is no longer owned solely by IT or the business. It needs to be a collaboration that ensures everyone gets the capabilities and control they need. Yet I often see organizations tackling self-service projects without proper representation from IT and the business.
If BI is solely IT-led, business users will be disappointed in the capabilities they are given. If it’s solely business-led, IT gets upset that the business is operating outside established security and governance protocols. To make matters worse, even when the business tries to adopt a tool unilaterally, IT usually ends up getting involved down the road to connect to certain data sources or provide custom views for dashboards and report.
For a successful data analytics implementation, organizations need to create a balanced deployment team that leverages the technical strengths of the IT team with the business knowledge and requirements from the business side.
Mistake #5: Lack of Training
Self-service tools are supposed to be intuitive, but they’re not autonomous. One of the biggest mistakes an organization can make in their data analytics implementation is not investing in training. Empowering self-service users starts with adequate training, particularly since many of these users will be new to BI toolsets.
The good news is, according to the 2015 State of Self-Service BI Report, the top area for self-service investment in the next 12 to 24 months is end-user training. By shifting focus to improving the people and processes that power BI, IT can work around limited budgets and still address gaps in user skill sets, both of which are restraints to self-service BI adoption.
Adoption and success in data analytics is never going to happen overnight. But if organizations can address these five common mistakes from the beginning, they will notice an increase in user adoption of their data analytics tools—not to mention happier, more data-driven end users.
(About the author: Sana Narani is operations manager at Logi Analytics)