Increasingly, vendors are talking up self-service business intelligence tools for analyzing and visualizing data. The term self-service might imply that business users can do everything themselves and don't need help from IT and BI teams to gain valuable insights from business information. As with most things, though, reality lies somewhere in the middle.
Let's start by reviewing a little history. I began in this business in 1980. Back then, when a knowledge worker wanted to understand business data, an IT analyst would interview him; the analyst then would create documentation and pass it to a programmer, who would place the request in his work queue. Eventually, the required coding would get done, and reports would begin to appear in printed form in the business user's mailbox or possibly on a very simple, mainframe-based CICS screen. Hopefully, the reporting requirements hadn't changed in the meantime.
Along came the PC, and individual business departments started to bypass IT and create their own little databases in Excel -- and then in more powerful tools -- enabling them to build their own reports. The resulting decentralization empowered business users, but data analysis capabilities remained pretty rudimentary; it also made effective data governance a challenge and maintaining a consistent business model across an organization almost impossible.
Then came enterprise resource planning, customer relationship management and sales force automation systems, followed by the first generation of BI vendors, offering products focused on integrating and reporting on data stored in different systems and applications. Often, though, those early tools were difficult to use. Building BI reports required business analysts and programmers in IT to repeat the joy of the early 1980s.
Riding the data warehouse wave
The 1990s brought the rapid growth of companies such as Business Objects and Cognos that rode the wave of data warehouses and data marts while not limiting themselves to those sources. The BI software packages they delivered were the first tools with strong functionality that allowed business users to create and manipulate reports and charts based on complex data. IT had to install and maintain the BI systems, and business analysts were still needed to link data sources, but it was a huge leap forward -- and it led to the concept of self-service BI tools.
As more user-driven analysis and reporting became possible, it's no surprise that additional functionality was requested. Advances in hardware sped processing and lowered costs, enabling the development of software that supported more complex analytics and ever richer visualizations. A new generation of BI vendors -- Tableau and QlikTech being the most visible -- emerged with more powerful front-end tools. Business users could quickly find data sources and link them, then create their own reports, charts and graphics and pop them into dashboards they could share with other workers.
At least, that's what the demoware shows. But as much as vendors talk about self-service, strong IT involvement is still required. The user interfaces in self-service business intelligence tools haven't reached the ease-of-use level necessary for knowledge workers to do more than the bare minimum themselves. And dealing with the complexity of pulling together information from various data sources takes knowledge and core technical skills that most business users, and even many business analysts, don't have.
Data governance rears its head
In addition, data governance -- often ignored in BI initiatives -- is returning to the fore. There are far more government regulations and compliance reporting requirements to contend with now, and the advent of mobile devices has added new concerns. The need to address governance issues will lead to more security and privacy protections, which might limit the ability of business users to access data -- and, in turn, limit the notion of self-service BI.
So self-service doesn't mean that users can run BI and data visualization applications without any IT involvement or intervention. What is changing is that they're more able to analyze data and share the results of the analysis in a collaborative and visually engaging way -- but BI will never be as simple as the retail model of self-service makes it sound.
Self-service BI deployments need to walk a tight line between data access and data discovery on one side and data governance on the other. More powerful analytics and visualization capabilities will continue to become available to business users -- but just as you don't refine the gasoline that fuels your car, self-service doesn't mean taking over the data supply chain. Users of self-service BI tools should think not about gaining independence from IT and BI teams, but about developing a better working relationship that will make it easier to analyze data faster and more accurately.
About the author:
David A. Teich is principal consultant at Teich Communications, a technology consulting and marketing services company. Teich has more than three decades of experience in the business of technology. Email him at firstname.lastname@example.org.
Self-service BI doesn’t mean 'one size fits all'
Will data scientists become extinct thanks to self-service BI tools?
The benefits and challenges of mixing self-service BI and CRM