For some businesses and some situations Do-It-Yourself Business Intelligence (DIY BI) provides a real opportunity to drag key business information and analytics, kicking and screaming out of their company’s data systems. The notion that the end user can make decisions based on the quality of their own analytics, using easy to use software tools to create ad-hoc reports and visualisations, is both empowering and laudable.
However, the vision of a non-technical employee or business owner firing up an analytics package, connecting to one or more data sources, and using what is usually a drag and drop interface to build the complex dashboards and reports needed for their organisations, is in our view, a vision too far.
Choosing the right DIY BI software is key in this process. Some DIY BI programs present themselves as self-service tools, but require a significant understanding of data connectivity, measures, dimensions, queries, parameters, drills and filters, as well as the broad technical language of data analytics and the configuration of access and security, so as to render them almost unusable in a self-service environment. Or at best they present the user with development and learning challenges, which only the most dogged will overcome.
Other packages use a range of easy to configure widgets that can connect to pre-defined data sources, as long as the user has the right login credentials. These packages come pre-populated with some basic, and often useful, data analysis, but the analysis that is available is usually pre-defined and generally does not provide some of the key business specific measures and comparators that are often required, or that the end user actually needs in order to inform actions and to move their business forward.
DIY BI has its place, but BI development in the hands of specialists can open up a wealth of potential and opportunity, and can be both beneficial and cost effective.
Is using a BI specialist more expensive? Arguably, yes it is. Paying someone to build and maintain your data analytics platform may well be seen as more costly than doing it yourself, but often doing it yourself does not factor in the time taken by the non-specialist to build a useful and maintainable system, and in many cases, having spent that time, people often do not get the reports and visualisations they would like, because of their inability to fully develop their BI platform.
Putting a company’s BI development in the hands of specialists has often been viewed as a process of taking BI away from the end user, allowing the specialist to impose a set of business measures on them. This has not been my experience. Working closely with the end user as both a developer and a consultant, it is possible to put into place systems that provide easy access to information that is informed by the end user, but which is also built upon by developers who have experiences of other presentation ideas and metrics that they can bring to the party.
Instead of seeing IT specialists as the gatekeepers of access to BI who have to be consulted periodically to provide the information that an organisation needs, they should be seen as the data specialists who can open up a whole wealth of relevant information streams that are informed by business needs in consultation with owners, managers and other information stakeholders.