Increasing Relevance of Self Service BI
Self-Service BI : Power of BI to create reports without having dependency on IT, reduce latency, when and how business wants to create it.
Key Reasons for Self-Service BI demand : Rigid BI processes to get the data from source to end reporting business consumers via complex IT processes, thus creating dependency, latency and eventually business users losing the opportunity to discover, take decisions when it matters most. That's the business view of the problem, however, the IT view of problem is about governance, control of data, redundancy, siloed data sets/reports, data security, access controls & inconsistency across enterprise. The CIO's are constantly struggling to have the right balance between the two concerns.
Paradigm of Self-Service BI : Since Self-service BI is all about business consumers of Information, there's a need to understand how business leverages the Information & the core differentiation of the business consumers defining their profiles. Although both can be ad-hoc, the appetite for Information is quite different.
1. Report creators - Some terminologies put it as ad-hoc report analysts, power users or data discoverers. Such users are capable of writing SQL Scripts on data, transform datasets, aggregate, build new data relationships, data profiling and enriching datasets beyond creating new ad-hoc reports for second category of business consumers. These are demanding users needing access to various data sets to discover patterns & do data analysis before designing consumable reports.
2. Report consumers - Executives, Departmental Heads, Top management, Casual users primarily interested in viewing the defined existing reports. Actions such users typically perform being Slicing, dicing, filtering, summarizing, charting/graphing, aggregating, exporting to Excel or other common formats convenient to them.
Challenge for IT : It's the report creators or power users which pose bigger challenge for IT in terms of meeting their constantly changing demand for data. The situation can easily run out of controls if not governed via well defined and business inclusive framework, as power users might end up creating siloed reports, views which are never consumed eventually or worse creating a long running cross-join queries on source impacting it's performance. The loop of business funded BI projects, and IT running the show only adds to this challenge thus becomes critical to have Governance board which constantly reviews, defined policies, standards and process of how Self-service reporting will be published to larger user base. Such a governance board need to have representation from IT and Business alike, and business power users working closely with IT on their constant changing needs yet aligned to business groups. On the other hand IT will require to have data analysts as part of their team to work with business power users providing adequate trainings, tools experience/adoption and keep pace with the changing business landscape.
How to address this challenge : A well defined Governance, and a framework which is based on the right classification of users will help identify the needs at each level, when, in what format, how and various tools to be leveraged for each user base.
What to expect from tools to cover for a self-service BI? The diagram below depicts the various aspects that one should expect BI tools to cover, it may not always be 1-tool-fits-all solution as it will differ with the user classification and needs. A Power user needs a report designer tool which can query source DW/EDW or Marts, create data views, joins, transformations before deciding to create a viewable report, whereas a casual user or executive level user might only needs report viewing tools which allows operations like alerts, slice-dice, charting, drill down/up/thru or filter capabilities.
Figure 1 - Self-Service BI Capabilities
Self-service BI market has recently been flooded with some very mature tools like QlikView, Tableau, Tibco Spotfire, MS Power View to name a few & each of those tools have one thing in common, they are quite easy to use and business users get to use those tools on their turf e.g. Excel based add-ins is common to most tools. The reference provided below gives a good insight on various tools that are competing in this space. The best way to decide your tool(s) is to do a tool evaluation exercise by running thru various features of shortlisted tools in market, & allowing the tool vendors to walk your IT/Business stakeholders with demos, limitations, trainings, enablement and ad-hoc capabilities they bring on table. However, before one embarks on the tool evaluation journey, it's critical to define self-service BI in context of your enterprise, expectations from self-service BI as capabilities to accomplish business goals/KPI's and classify the various business information needs into business users.
1. Analyzing Gartner's 2014 Magic Quadrant for BI and Analytics Platforms - http://www.jenunderwood.com/2014/03/16/analyzing-gartners-2014-magic-quadrant-for-bi-and-analytics-platforms/