This post is part of a series on trends that highlight the two worlds of BI: the world where, according to industry pundits and vendors, enlightened companies have achieved BI nirvana, versus the real world, where business and IT folks are simply doing the best they can. Let’s get realistic about what most people are doing and how to achieve this elusive nirvana.

Which of these scenarios applies to your enterprise?

1. Business people are routinely using analytic-driven processes to manage and grow the business.   Advanced analytics tools such as data visualization, predictive analytics and data discovery are being used pervasively across the enterprise. Business people are using consistent, clean, comprehensive and current data in the analysis that is used in decision-making. They have achieved analytical nirvana.

2. Many business people are primarily using enterprise-wide BI applications, typically dashboards and reports, for the analysis they need to make decisions in their jobs. Their data is consistent, clean and current.

3. Most business people are using a mix of the enterprise-wide BI applications and reporting supplied by (and specific to) the various business applications. The business people generally analyze data in reporting silos either because they do not need an enterprise-wide perspective or they do not trust what they see because of data inconsistency.

4. Some business people are using enterprise-wide BI and/or application-specific reporting. But mostly the company is using spreadsheets to gather, integrate and report on data from data silos. They spend much more time gathering data and then reconciling differences between different reports than analyzing the data. Because data is not consistent, clean and current, they make business decisions on the best available data or, simply on the report that reinforces the business person’s own narrative.

The case studies, white papers and articles that you see in the industry press make it sound as if everyone is at least in scenario 2 and many enterprises achieve scenario 1 – analytical nirvana. If you believe that, you could get an inferiority complex, because you know your own company has a long way to go.

The real deal

The truth is that even in companies that are touted in case studies for their terrific enterprise-wide and self-service BI efforts, you’ll probably find many data shadow systems or spreadmarts. Business people there will still be using spreadsheets as the “final” BI tool.

Tom Davenport’s excellent article Analytics 3.0 in the December 2013 Harvard Business Review discusses the evolution from Analytics 1.0 to Analytics 3.0 (real-time, predictive, pervasive and the cornerstone of decision-making.) But even in firms with huge investments and pockets of Analytics 2.0 or 3.0, you would probably still find Analytics 0.5, i.e. the spreadsheet, being the most pervasive business analysis tool.

So maybe you shouldn’t have an inferiority complex, since so many other companies are lagging too.

Moving to an analytical nirvana

Enterprises need to strive to improve and move towards this analytical nirvana. The improvements they need to get there, however, may be more foundational, and certainly more work than simply buying the latest and greatest advanced analytics tool. These advanced tools are probably quite beneficial (to a select group of business people), but enterprises need to understand that they are not the silver bullet to nirvana.

Three of the foundational changes companies need to make to achieve analytical nirvana are:

Get serious about data governance.  Data governance, both for data definitions and business agreed-upon metrics, is foundational because if data is not consistent, then slick visualizations do not matter. The old school term was GIGO (garbage in, garbage out).  I am not saying everything has to be perfect for BI to be useful, but if business people spend their time debating the numbers or reconciling data (because they do not trust it), then foundational work needs to be done. It’s not glamorous, but it sure is essential.

Get serious about data integration. Most enterprises have a data integration backlog, but keep getting “distracted” by industry pundits and vendors that claim that this time, with the latest and greatest analytic tools, they do not need to integrate (gather, cleanse, standardize, conform and transform into multiple business processes) data but can simply “point & click” to get their answers. If it only were that easy…

Get serious about spreadsheets. What’s wrong with having the “lowly” spreadsheet as part of an enterprise analytic portfolio? Nothing. The problem with data shadow systems is not that the spreadsheet is being used for reporting, but that it is being used as the data integration and transformation tool without any governance or architecture. If a spreadsheet is used to analyze the same consistent, comprehensive, clean and current data as the dashboard, data discovery and data visualization applications then it is a viable tool.  Many business people gravitate to it since it is truly pervasive. I understand that many IT folks will cringe at this recommendation, but have they seen what can be analytically done with spreadsheets nowadays, especially if they are leveraging an information backbone?

Next steps

Enterprises will obtain terrific business value from advanced analytics if they have the data available for business people to use with these tools. It may be great to purchase and use these tools for specific business situations, but do not forget that the foundational work is needed for these solutions to become pervasive. Then maybe you can be written up for a case study too.