Advanced Analytics Article “Moving from BI to advanced analytics” Review

Article authors were  Scott Vanvalkenburgh (SAS) and Bill Franks (Teradata) - published in SASCOM 2011 for second quarter.

“Standard BI has many advantages but is very limited ability to predict, forecast and make inferences. It has limited functionalities in finding hidden patterns and projections in internal data like customer behavior, fraud, likeliness to churn and similar. Data mining and advanced analytics are designed and structured for this advanced features.”

“…creation of consistent data, rigorous system governance, and sophisticated data integration and data quality can serve as a sound foundation for advanced analytics.”

Article in general emphasizes usage of advanced analytics with sandbox technologies with good notice that standard Business Intelligence solutions can not solve tremendous analytical bourdons faced in front of business analytics. Generally, I do like promotion of new technologies but there are pitfalls behind endorsements of advanced analytics and caution is needed before accepting them in corporate Information Systems.

Advanced analytic solutions will not solve financial implications of analytical results such as customer profitabilities. I disagree with subtle statement that new insides into business logic will automatically raise business value. 

Advanced analytics will add new knowledge to decision makers. New knowledge will not cancel and erase old knowledge, it is only additional knowledge. Here are some thoughts and questions.

• Does additional knowledge has direct relationship with financial statements?
• Any partial relation in sandboxes of advanced EDW solutions (Teradata) with financial data will bring coarse error and advanced analytical results will not be useful any more.
• Advanced knowledge can create more information overload then ever if it is not structured in complete information picture of company.
• “Absence of thorough analysis can lead to inconsistent conclusions or even operational decisions that are just plain wrong”
• Does thorough analysis includes relations with other analysis. Surplus in one analysis usually means minus in other analysis. Resources are limited.
• Does thorough analysis include financial contribution and is it shown within official financial statements?

Advanced analysis can usually be just another set of data, someone’s alibi for doing some work or real value. All depends how it is used in complete information flow.

Who needs more knowledge in situation when even current knowledge islands are not properly integrated? There has many talks about information integration and this is new mantra for vendors of Master Data Management solutions.

Well,

knowledge integration is something completely different then data integration. It is not same and will never be same.

“…to be able to calculate profitability of customers…” Yeah, this is very brave wording in article. Profitability is usually presented as one side of the game, as revenue generators and as very narrow cost generators…

Where are all costs and where is mother lord of all profitability Profit and Loss statement? Does this profitability shows contribution in PL? Very, very, very hardly!

So, what is actually named profitability in the article? More likely it is local wording for revenue views in this advanced solutions. Profitability, it is word reserved for financial gurus, not for business advanced analytics.

Advanced analytical users are usually business analytics. How will they calculate profitability? Who will prepare them financial figures and relations to their business figures? This is the task of controllers and not of advanced analytics.