Working outside of the IT org as we know it. This is a comment on the ability of traditional IT to support advanced analytics.

Edward H. Vandenberg

EMC IT Proven

lena2By Dr. Lena Tenenboim-Chekina — Senior Data Scientist, EMC IT

Smart data visualization is proving to be an essential tool in maintaining increasingly complex Big Data systems in the cloud.

The adoption of Big Data tools and technology heavily relies on distributed scaled out computing. One of the main differences in this setting is that it includes systems that operate as a whole on top of several independent hosts. These hosts coordinate their actions with limited information and as a result maintenance complexity significantly increases. One way to overcome this challenge is smart data visualization, which helps the IT experts and management pinpoint the source of problems quickly.

The need for smart visualization is not unique to this problem. Representing complex data as a concise picture which tells decision-makers a story is a key part of any data analytics or data science project. Valuable results of a rigorous analysis may…

View original post 1,250 more words

Who has it in your org today? Will they listen to the data science?

Your stakeholders lack a shared understanding of the methods and practice of advanced analytics.  You start out with a trust deficit when explaining how the mathematics will improve business results.

To build trust, start in advance by building an ordinary business relationship to the operations management.  Next share stories of analytics successes and how they were achieved (ideally those that you have directed). Next coach your stakeholders to interpret model results by simplifying the complex model validation process.

Gradually build an Arena of shared understanding for how models can help operations arrive at a better performance state.  This is hard work and not the stuff of algorithms and data but almost as important.

Look at the Jahari Window for expanding the Arena of trust.

Edward H. Vandenberg

In customer service environments, the best approach to analytics is to create models that engage the insights of people who are providing the service and engaging with your customers.  That’s a real challenge.  Most modeling today is to drive procedures and replace or enhance procedure logic in information systems.

‘Service Analytics’ enables superior human intelligence leveraged by powerful machine learning.  Some refer to these models as a pair of glasses to improve insight. Procedural models drives another generation of logical machines (rules engines) that have inherent limitations in service environments.

I’ve learned from Gary Klein.  Senior Scientist at MIcroCognition LLC “instrumental in founding the field of naturalistic decision making”.

Read “Thinking,, The New Science of Decision-Making, Problem-Solving and Prediction” edited by John Brockman.  Harper Perenial 2013.

Thanks Gary.

Edward H. Vandenberg

Interpretation of models must overcome the limitations of most people involving probabilities.  Natural Frequencies are the best way to present technical findings.  Please refer to:

“Thinking” ed John Brockman, Chapture 3 Smart Heuristics. By Gerg Gigerenzer.Introduction by John Brockman.

Gerd Gigerenzer – Wikipedia, the free encyclopedia.

Over the next few years, advanced analytics will emerge as a separate corporate function in some leading companies.  The activity is a mix of IT, Finance, Process Management and Knowledge Management that simply defies the current corporate functional matrix.  Add to that a highly skilled work force whose ethical behavior is (will be) a matter of corporate ethics and law.  It is a strange prediction, admittedly.   But Finance, HR, and Accounting started somewhere.  And that is where advanced analytics is now.

Astute Analytics Executives will position themselves for this evolution.  It is a positive development for the firm, what ever industry you work in.  Analytics is the customer interface in a mass market environment where the preferences of one is the required service or product model.

Don’t expect the new VP of Advanced Analytics to do it all.  When processes and decision-making have to change to exploit insights from data, the line managers, with executive sponsorship, must carry the weight.  Without that, you have models that work on paper but aren’t used in the operations.  It’s tempting to blame the analytics leader but that is misplaced.  Operations managers actually resist doing things in a new way, despite the math telling them otherwise. And the VP of Analytics likely has no power to change that, short of appealing to the executive staff–not a popular move for the analytics lead that also must evangelize for new projects and problems to solve.

Also, commanding a small specialized team is not normally seen as a position of clout, despite the title and sponsorship.  Managers of large operations (financial and headcount) naturally have more influence in most organizations, even if they have a peer title to the advanced analytics leader.

In defense of the operations level manager, they are rightfully reluctant to be accountable for mathematics that is probablistic and hard to understand.  Likely they are also normally protective over the operational influence that analytics can have within their business units.  This creates tension that makes analytics ROI go sideways.

What’s the answer?  For analytics to truly be exploited, operations management must step up….understand the science more, be ready to believe in it and lead their operations to adopting it.  That means they are hired for it, trained for it and managed for it.  How an organization mobilizes for that is under the leadership of a Chief Analytics Officer and a full program management approach to advanced analytics.  Deploying advanced analytics must be seen as the path to promotion for career operations managers.

Secondly, the advanced analytics effort must include Test and Learn experiments for every model pilot that help prove the in-use value of models beyond validation on historic data.  This is a natural extension of the model development work

Cognitive Decision Science must be part of your modeling solutions. Understanding how people make decisions and what their typical biases are is critical to developing decision support models.

Daniel Kahneman,. ” Thinking Fast and Slow’ will provide some guidance on this subject.

The types of decisions follow data types; each with their own defect modes. The best models inoculate the business from decision bias that is always present when people make decisions. If you are working in a ‘flow’ transaction environment, making good decisions, consistently at speed simply does not happen without data analytics support.

Share things that can (must) be shared

  • Specialized Talent
  • Infrastructure
  • Some datasets
  • Tools

 

Focus services to deliver on demand

  • Domain knowledge
  • Data and systems expertise
  • Capacity for high demand customers

 

Standardize things that will help deliver consistent quality

  • Methodology
  • Project Practices
  • Role Descriptions

 

Give synergy to the effort

  • Complimentary skills and knowledge
  • Knowledge sharing and imagination
  • Contrarian viewpoints

 

Control Risks Formally

  • Skills definition
  • Independent Quality Review of the models and interim work products
  • Checks on conflict of interests and influence
  • Management accountability (project level)
  • Sign-off and approval process
  • Ethical Standards

 

Develop Enterprise Assets

  • Reusable datasets
  • Documented models

 

An Enterprise Identity and Voice

  • Organizational voice
  • A place for people with unique skills to belong
  • Promote identity, value and scope of the work
  • Tell the story to the enterprise