It’s clear that there is considerable recent market movement towards self-service business intelligence (SSBI) in the numerous vendor offerings available. There is also a growing concern among the Data Science community that ordinary business users may misunderstand or misinterpret the available data, leading to incorrect results.

Experienced data scientists have a tremendous ability to analyze, compare, drill into, and view data in a manner that enables insightful market intelligence. Without that deep knowledge and wide exposure to business data, ordinary business users may misread the data and miss key insights.

Thus, though self-service business intelligence platforms are easily available now, experienced data technologists may still be required to aid business users in deriving strategic intelligence and then deliver them through friendly data visualization tools. Self-service BI involves far more than a suite of easily available and accessible tool sets.

In the post An Introduction to Self-Service Business Intelligence, the author explains how an average user can “filter, group, or segment” data for Analytics purposes without any technical knowledge of business intelligence systems, which was required in traditional BI.  The author also warns that as “one size does not fit all” in the SSBI world, the IT team must closely examine the needs, expectations, and skills of the users in question before designing a self-service business intelligence platform. The small margin of super-users still may need ad-hoc tools to conduct personalized analytics, therefore self-service BI systems have to be flexible and scalable enough to support a wide range of users.

Bernard Marr, indicates in Why We Must Rethink Self-Service BI, Analytics and Reporting that the danger of SSBI is making highly sophisticated tools available to novices. While over reliance on data technology teams is undesirable in the long run, enough caution must be taken to ensure that some key technical personnel must lead the business users at all levels to ask the right questions about their data and use the right tools to conduct the analytics for desired results.

Users should not be left alone with highly sensitive datasets or technical tools without proper training. The DATAVERSITY^® article Self-Service Business Intelligence is Big, but is it for Everyone? explores the possibility of transitioning into self-service BI from a traditional BI landscape.

Self-Service BI: Enabling Informed Decisions

An inherent capability of a good, self-service business intelligence platform is the ability to gather and compile multi-structured data from disparate sources and then convert that data into actionable intelligence. This trend in BI technology design indicates a gradual shift from a highly controlled, IT-led activity to a mainstream business activity managed by ordinary business users.

Powerful BI platforms usually have most of the following characteristics:

• The presence of good data preparation tools to overcome security and Data Governance risks
• The capability for handling big data for data discovery
• Open access to external data for generating instant reports based on need
• Simplicity of usage tools for conducting analytics and generating reports
• Scalability to include simple to complex analytical models
• Powerful visualization tools for viewing the results.

The article Self-Service BI: Empowering Managers to Make Better Decisions discusses how SSBI serves the corporate managers in times of needs.

Traditional Business Intelligence vs. Self-Service Business Intelligence

While the market value of traditional BI often came from its “branding,” backed by industry experts such as Oracle, IBM Cognos, or SAP Business Objects, a significant triumph of self-service BI over traditional BI is that is enables data discovery, thus providing a quick and easy channel to work around the “hurdles of data extraction & data preparation.” It’s necessary to understand how the data preparation tools in self-service BI aids business users to quickly gather, organize, clean, and prepare data from disparate data sources.

With Data Discovery, users can pull multi-structured data from multiple sources within an organization, and sometimes, from external sources too. The post Self-Service BI vs Traditional Business Intelligence makes an interesting comparison between the two. Here are some other comparative features that distinguish traditional BI from self-service BI:

• Traditional BI, because of its branded footprint, often costs more and is harder to implement. Self-service BI is relatively inexpensive in comparison and easier to implement.
• Self-service BI supports a wide variety of data sources, while traditional BI usually relies on data warehouses.
• Traditional BI requires some programming or SQL knowledge, while many self-service business intelligence users do not know how to code.
• In traditional BI, ordinary business users have to depend on data analysts for complex analytics or high-end reports, while in SSBI, the end user is perfectly capable of conducting analytics or generating reports without any technical help.
• While traditional BI was once the guarded domain of financially solvent businesses, SSBI helps small businesses with low capital to implement in-house BI without much financial commitment.

A survey conducted on the state of Analytics Adoption, conducted by Impact Analytix for embedded BI vendor Logi Analytics, reveals that more than 65 percent of the survey respondents have already switched over to self-service analytics solutions.

Can Self-Service BI and Self-Service Analytics Replace Data Scientists?

The goal of self-service BI is to empower the business users to find their own actionable solutions through the use of a guided analytics platform without the presence of a data technologist. However, the Forbes post Why Self-Service Analytics Won’t Replace Data Analytics Professionals, May Help Them reveals that this particular goal is still at a distance, with technologies like OLAP or data discovery only meeting partial user objectives. On the one hand, the users want complete freedom from technical experts, on the other hand, they are still not on a comfort zone with advanced self-service BI technologies and tools.

In the case of data discovery, users have often displayed a fascination for dazzling graphics, while the essentials of deriving meaningful insights from data patterns may have been undermined. Despite the easy availability of “modern self-service BI tools,” most business users preferred remaining with one business application instead of constantly switching to a separate self-service analytics platform.

Users want the power to quickly extract and combine data from disparate sources, and then explore and query the data to arrive at instant business insights to enable prompt decision management. Technologies like big data and Hadoop have made it possible to some degree, but interpreting combining and complex data to detect actionable insights is still a long way off.

How to Enable Self-Service Business Intelligence

New technology challenges emerging with continuous advancements require a thorough understanding of the entire Data Management ecosystem comprised of big data, Hadoop, data discovery, data visualization, and other related technologies.

The article The Hidden Costs of Self-Service BI Initiatives hints at a crucial operational concern for Self-Service BI professionals. As different data analysts may work independently at different times, redundancy in dataset modeling is likely to occur, thus increasing the overhead cost of SSBI. The concurrent execution of the same datasets with the same results will also exhaust system resources. On the other hand, analytics completed for specific needs or for a small audience can generate reports that are hard to digest and increase the cost. When an individual develops data models, and some other person or team generates reports, this problem can frequently happen. Thus, a far better approach may be to clearly articulate assumptions about data preparation, data modeling, and Reports during the initial stage.

The Gartner article How to Enable Self-Service Analytics and Business Intelligence: Lessons from Gartner Award Finalists has a wealth of lessons, which can be immediately put into practice by self-service BI professionals:

Some of those lessons include:

• The desirable goals or business outcomes of SSBI should be aligned with organizational goals and expressed in terms of measurable benefits
• The central IT team and ordinary business users must work together through the different phases of SSBI system design, development, and usage support
• The business users’ own innovations and contributions must be supported by taking a flexible stance on data and analytics governance
• Equip the business users with a well-drafted “Engagement Plan” to ensure Analytics success on the self-service platform

Self-Service BI: Data Security & Data Governance Issues

Businesses contemplating self-service BI implementation must consider engaging data security and Data Governance experts to ensure that a new self-service BI implementation goes through pre- and post-implementation security checks.

Data security on self-service BI platforms are important for many reasons, including: Digital privacy measures to prevent unauthorized access to data, freedom from data leaks, prevention of internal misuse, and human errors in data use.  The article Delivering Governed Self-Service BI across the Enterprise throws some light on ensuring proper governance in enterprise-wide Self-Service BI. The article includes a Data Governance checklist.

Some self-service BI platforms are still lacking in proper data security and Data Governance controls. Well-designed self-service BI platforms usually have data preparation tools that store, manage, and provide access to source data, prepared data, and data models with appropriate governance measures without hindering the self-service analytics processes.

A significant number of self-service business intelligence solutions offer advanced Data Governance capabilities like data masking, data retention, data lineage, role-based access, and auditing. Data Quality and Data Governance are so important for self-service BI to succeed. Even with the best-of-breed visualization tools, an SSBI system may fail to deliver results if the quality of the data is not sound.

Image used under license from Shutterstock.com