Business intelligence has a long history. Historian Richard Millar Devens first used the term to describe the machinations of banker Sir Henry Furnese, who collected information and acted on it quickly to outsmart his competition. Today, the term describes that same activity, but on a much larger scale, as organizations race to collect, analyze, and act on data first.
But there have always been limits on who can access valuable data, as well as how it can be used. With remote and hybrid work on the rise, the ability to locate and leverage data and expertise — wherever it resides — is more critical than ever.
A Short History of Business Intelligence
Modern business intelligence starts at the dawn of digital computing. In the 1970s, data was confined to mainframes and primitive databases. Reports required a formal request of the few who could access that data. The 1980s ushered in the antithesis of this version of computing — personal computing and distributed database management — but also introduced duplicated data and enterprise data silos.
During the 1990s, attempts were made to tackle challenges including:
- Inefficient data silos.
- The expense and time required to create reports from transactional data sources including mainframes and minicomputers
- The impact on the performance of transactional applications at the time due to the number of reports being created
This led to the birth of separate systems for reporting: the enterprise data warehouse. For the first time, the focus of a system became business questions, where data was denormalized. But given the nature of organizations at the time, the focus was upon creating executive information systems (EIS). (One of our authors participated in one such project during that period.)
A shift emerged around 2000 with the initial discussions regarding digital transformation. Organizations began considering information democracy and questioning the cycle time required from report creation to decision making. The request model started to fray. Slow requirements led technology leaders to demand proactive business intelligence.
As Business Objects founder Bernard Liautaud notes in e-Business Intelligence: Turning Information Into Knowledge Into Profit (McGraw-Hill, 2001), the lack of ad hoc data access causes IT staff to drown in requests. When business users design their own queries and reports, programmers can focus on their priorities and high-value projects. “In a changing economy that places a premium on speed of execution,” Liautaud writes, “it is critical to reduce the cycle time of decision.”
The Failure of Traditional BI Approach
Traditional BI approaches and technologies — even when using the latest technology, best practices, and architectures — almost always have a serious side effect: a constant backlog of BI requests. This happened for many reasons.
First, data and analytics teams never were comfortable ceding control up to business teams. They weren’t confident that business users could make sense of huge data sets, and felt that even business users who were trained in BI tools didn’t use them often enough to retain their skills.
As a result, “super users” and business analysts emerged. These so-called “citizen data scientists” remained a roadblock between business users and data — and between data and decision making. Business teams still had to request data. Although it became easier for BI and analytics teams to create custom reports and dashboards in tools such as Tableau, Looker, and Power BI those tools still isolated the user from data. Former CIO Isaac Sacolick reflects on this data-inefficient past: “Remember the days when reporting was centralized in IT? Days/weeks old, backward-looking reports that no one used? To be data-driven means everyone should ask questions and have access to tools and data to pursue answers.”
The Emergence of Self-Service BI
Tech Target editor Craig Stedman, declared self-service business intelligence “an approach to data analytics that enables business users to access and explore data sets even if they don’t have a background in BI or related functions like data mining and statistical analysis.”
At Alation, we believe self-service has three unique stakeholders:
- End users trying to discover data for decision making
- Business analysts needing to find data to create new analysis and reports
- Data scientists needing to quickly find and profile data
Self-service BI requires carefully labeled data and the ability for business users to access and explore data. CIOs and CDOs want to reduce the cycle time for decision making, and make data and analytics teams less of a bottleneck for businesses needing to respond at the speed of today’s business.
This requires fundamental change:
- Business users must get the information they need, when they need it — no more response backlogs!
- Business analysts must respond faster. This is accomplished by business users self-discovering data and by eliminating duplicate data requests that slowed the pace of decision making.
To be clear, the corporate goal is to make the right decisions when they are needed, and this requires quick and easy access to the right data at the right time. “The speed of business and competition now requires that stakeholders have quick access to meaningful data to make decisions,” says Net Health CIO Jason James regarding the drive to self-service. “Data needs to be up-to-date, relevant, and relatively easy to access.”
That’s why the value of self-service centers around changing how organizations think about what they’re doing and why. CIO David Seidl suggests, “A highly usable data tool including data discovery and contextualization is critical for more casual, non-power users,” suggests Miami University CIO David Seidl. “That’s the real destination of self-service BI in the long term, with a stop at power users along the way — as well as discovering you don’t have the data (or that it’s not complete, bad, or un-governed) and figuring out why, and if it’s worth gathering and maintaining too, right?”
It’s also important for users to discover which data they should use for decision making. “You need a way to build data confidence,” says analyst Jack Gold. “Users need to assess how good data is. Next, they need to make that data accessible (securely through governance) so people can access and use it. Finally, you need simple-to-use discovery and analytics tools that don’t need programmers.”
The primary goal of moving from legacy to self-service BI is to drive better-informed decision-making to achieve positive business outcomes including increased efficiency, better customer satisfaction, and higher revenue and profits. This is accomplished by making data and reports discoverable by business by doing something users already know, a google like search. This is why author and consultant Geoffrey Moore famously argued that software move from systems of record to systems of engagement.
In this approach — in contrast to business intelligence — users request analytical queries only for new data topics. This frees the BI analyst or other BI professional to focus on unique requests.
To learn more about self-service, we recommend:
Remember that “self-service” encompasses many related things, including the ability to create analytics faster. But at its core, it’s about making data discoverable, trustworthy, and shareable. This makes everyone more productive and businesses able to outpace both demand and the competition.
Have questions? See the data dialogs on self-service to learn more: