Shadow BI: shady or open for business?

Shadow BI is a common phenomenon in any organisation where the business has an Open or Microsoft Office on the PC; i.e. 99.9%  of the users can mash up data in spreadsheets, perform rudimentary descriptive and test statistics and some predictions using linear regression. Some of them download open source data science tools like Weka and KNIME and take it a step further using fancier regression techniques as well as machine learning and deep learning to come up with new insights.

On October 18, BA4All’s Analytic Insight 2016 had a peer exchange with about 60 analytics professionals to reflect on three questions:

• ·        What are the top three reasons for Shadow BI?
• ·        What are the top three opportunities Shadow BI may bring to the organisation?
• ·        What are the top three solutions for the issues it brings about?

The most quoted reasons for Shadow BI

Eww! IT is taking some heavy flak from the business: “ICT lacks innovation culture”, “IT wants to control too much!” and especially the time IT takes to deliver the analytics was high on the list.

Other, frequently mentioned reasons were the lack of business knowledge, changing requirements from the business  and the inadequate funding clearly indicate a troubled relationship between ICT and the business as the root cause for Shadow BI.

Yet, opportunities galore!

Shadow BI can improve efficiency in decision making provided the data quality is fit for purpose.  In case of bad data quality it may provoke some lessons learned for the business as they are the custodians of data quality.

This under-the-radar form of BI can also foster innovation as users are unrestrained in discovering new patterns, relationships and generate challenging insights. and provide faster response to business questions.

A mix of tech and HR came up in the discussions

The top 3 solutions for issues with Shadow BI

The group came up with both technical and predominantly organizational and HRM solutions. Here are the human factors:

• ·        market BI to the business and IT people,
• ·        governance (also a technical remedy if the tools are in place)
• ·        empowerment of the business
• ·        adopt a fail fast culture
• ·        knowledge sharing and documentation
• ·        strategy alignment
• ·        integrate analytical culture and competencies in the business
• ·        engage early in the development process

And these are the technical factors:

• ·        governance tools
• ·        Self-service BI and data wrangling tools
• ·        Sandboxes
• ·        Optimise applications for analytics

For a discussion on some of the arguments we refer to our next post in a few days