Either the truth is massaged by layers of staff delivering a sanitised pictures of customers, operations and the enterprise or too little data is analysed.

The latter is typical. it has been shown that over 80% of data in an insurance company is unstructured and inaccessible. No wonder decisions are made on incomplete knowledge.

Enterprise search is hampered by this vast amount of hidden, dark data.

See "Lighting up Dark Analytics"

Then there's the fact that enterprises and public sector organisations have just too many BI & Analytics tools lying around. It started with ERP solutions and then the addition of analytics apps like Business Objects, Cognos. Oracle, Micro Strategy and others.

Then line of business people, frustrated at the delay waiting for central IT to restructure inflexible metadata layers and interpret business needs wrongly, licensed analytics tools out of revenue budgets.

Visually exciting, low entry cost and with Qlik, Tableau and Yellowfin amongst others dashboards appeared and even the C-Suit using iPads to review insights. High on appeal and low on decision-making fact. People with inbuilt bias twisting the truth to support pet projects and ideas. Often unconscious bias but still hiding the truth from the CEO

See "Cognitive bias that distorts analysis"

What can be done about this?

One way forward- instead of islands of analytics drawing dathttp://www.analytics.world/post/102e3z0/embedded-analytics-drives-user-adoptiona from unconnected silos seek online platforms that can access and analyse the structured and unstructured data across an enterprise. That allows people to test ideas and suspicions across all the data. That gives the CEO the access to truth.

See 360 degree Insight 

Secondly, consider BI as an integral part of the apps and solutions that people use for their regular tasks and roles. Standalone BI and Analytics present insights out of context of the jobs people do. 

Far better to embed analytics in these apps and digital platforms so that a wider base of data is analysed and users see information within the context of the job in hand. 

See: "Embedded Analytics drives user adoption"