Dream to Learn is shutting down...

We are very sorry to say that Dream to Learn will be shutting down as of December 28th, 2019. If you have content that you wish to keep, you should make a copy of it before that date.


Got JARVIS? Executive Decision Support

POSTED IN: Data Analytics & Visualization Blog













Executive Decision Support: The $1 Trillion Dollar Question

A few weeks ago Fortune magazine released its annual ranking of top 500 US companies.  http://beta.fortune.com/fortune500

Big companies.  Big money. 

The combined revenue exceeds $12 trillion (two thirds of the US GDP), directly employing nearly 30 million people, and with combined profits of $840 billion!


A Thought Experiment – 10% Wiser

These are impressive profits that are powering the largest economy on the planet.  But what are the key factors that result in year after year of healthy profits at these well-capitalized and mature organizations?   

In addition to a favorable political, economic, social, and technological (i.e. infrastructure) environment – these organizations generally have good stewardship.  Good decision makers.  

Executive Leadership at each company must make thousands of decisions each year.  Most are small or medium sized operational decisions.  And a few decisions are much larger and very important – such as acquisitions or changes to corporate strategy.  Boards of Directors are also involved in key decisions around strategy and risk.

Now let’s consider a thought experiment:


WHAT IF each leader at each company had a ‘magic’ decision support mechanism that improved organizational decision making by 10% ?

What if they had access to a Corporate JARVIS?



For some organizations – improvements might take the form of better outcomes for numerous medium sized decisions.  For others – it may come down to a better informed stakeholder group and one or two critical strategic decisions each year.


Examples might include:

  • Deciding to withdraw from tantalizing acquisition or merger
  • Increasing investment in a regional or product expansion
  • Crisis Management – better situational awareness and less chance of groupthink
  • Workforce Changes – clearer view of global workforce and  
  • Terminating much loved but unprofitable areas of business. 
  • Evolving the company’s strategy after a boiling-frog moment


What impact would 10% improved decision making have on profitability?   

For Fortune 500 companies – how much would they value a system to help them avoid one large flub - each year?



ROI – Return on Intelligence


Ok – so the thought experiment where Fortune 500 leaders are 10% ‘smarter’ and made 10% better decisions - or reduced bad decisions by 10%.  More revenue, better use of capital, happier employees, better return for shareholders, lower risks. 

All is well in the land of Hypothetical - and the hypothetical shareholders rejoice!

But is this feasible in the real world?

Let’s explore how it might be possible by creating a better informed leadership team….


Decision Support Systems – Classic Model


There are three fundamental components of a traditional Decision Support System (DSS)

  1. the database (or knowledge base),
  2. the model (i.e., the decision context and user criteria)
  3. the user interface



Logical.  We have (1) data and knowledge – information corpora; We have a (2) models (sometimes a computer model, sometimes a “mental model”, usually an ensemble) and we have some form of (3) user interface to get at the system, interact, engage.


Whether you are Warren Buffet engaging with his team Denny’s – or the US Military with the finest computers and systems – the model is pretty standard.


Decision Support System 2.0 – Cognitive Catalyst


A Cognitive Decision Support System (CDSS) also called an Intelligent Decision Support System (IDSS) is a decision support system that makes extensive use of cognitive computing and artificial intelligence.


For many years there has been general agreement on the POTENTIAL value of a CDSS in business. 

But now we are now at a point where the technology components exist to build a useful and valuable system to help with decision support.   

Some early entrants such as http://equals3.ai/ are off to a great start and demonstrating how the pieces can fit together to deliver value. 

Cognitive can add value to each component by:

  1. the database – surfacing additional signal / information from unstructured data
  2. the model – helping users develop and test additional models. And bias reduction.
  3. the user interface – helping users access more data, more intuitively, more often

The next 5 years are going to be interesting in the C-Suite.  We now have the ability to make a CDSS a reality.   Nobody knows about the 10% number, but we can be certain that a better informed leadership team is going to make better decisions for the shareholders – and at Fortune 500 scale, substantial ROI is likely.

For the CDSS pioneers - let’s compare the approaches most likely to fail, with those most likely to succeed:

Top Ten Ways to Fail:

  1. Set expectations high.  Launch.  Fail.  Lose trust of executives.  Forever.
  2. Forget the end user experience and focus on the technology.
  3. Lots of Dashboards.
  4. Lots of big TV screens and technology.  And jargon – lots of technical jargon.
  5. Training sessions for executives so they can learn the system. 
  6. Let the consultants figure out what data you need.
  7. Expect ROI in first 18 months and stick to plan.
  8. Keep the project in the company.  Don't collaborate with other organizations.
  9. Hire consultants. Make assumptions about what executives need and want
  10.  Assign the project to one small department in one division.


Top Ten Ways to Succeed:

  1. Educate executives on the long journey ahead.  Build trust with straight talk.
  2. Think deeply about the challenges without tech. What is problem to be solved?
  3. Dialog! Humans have 100 billion neurons and 100,000 years of practice talking
  4. Let technology serve the mission. Discuss ideas as if talking to Warren Buffet.
  5. Design intuitive technology.   See: iPad in hands of a 3-year-old.
  6. Survey your data. Develop a deep understanding of institutional data & knowledge.
  7. Work with a five-year time horizon.  Be explorers.  Expect course corrections.
  8. Communicate with other companies.  Share ideas and mistakes.  Help each other.
  9. In house talent.   Few or no consultants. Executive sponsors for input and support
  10.  Create a small, full-time team from across the organization.   No silos.



Let’s Talk!   Dashboards Suck. 


Item #3 above is probably the most important one.


Dashboards generally suck.   Not always.  But more times than not they result in a “oh - that’s interesting” (not to be confused with “Ah! That’s helpful/actionable”)

Above: An Executive Dashboard that Sucks


We often build dashboards because we CAN build dashboards.  They are pretty.  They demo well.  While companies continue to buy them for ‘good fit’ jobs – there are also too many sold for jobs where the fit is poor.  And that’s bad.  The statistics on value delivered from IT investments support the assertion that many solutions are misrepresented.


Here are five reasons dashboards suck for many executive decision support use cases:

  1. Each middleman inserted between the data and user degrades the signal and loses nuanced information.  Analyst to Director to CEO?  You’re losing information. 
  2. Most CXO’s are communicate big ideas from their big brains using speech. And occasionally a white board to introduce or reinforce a shared mental model.
  3. Average age of CEO is over 50. They are busy. They don't want to click a mouse. They want to articulate the context, ask question(s) and get smarter.  Real time.
  4. If meeting participants are looking at a screen, they are not looking at each other.  Which means they are not communicating as effectively as possible.
  5. Dashboards are most effective when people already know what the dials communicate (e.g. like a car’s speedometer and tachometer);


Solution Components

  • LISTEN - services to listen to spoken word ( Speech To Text STT service)
  • INTERPRET - services to interpret spoken word (natural language understanding)
  • UNDERSTAND – architectures to weave it all together (sensemaking)
  • COGNITIVE – architectures to remember context; history;
  • EMPATHATIC – mechanisms to be more human;  to sense human emotions/tone
  • STRUCTURED DATA – aggregation, access, analysis
  • UNSTRUCTURED DATA – analysis, signal extraction, transformation, aggregation
  • VISUALIZE – when helpful, create and project visuals
  • PREDICT – mechanisms and algorithms to predict – and also explain conclusions
  • LEARN – system will make mistakes.  Architect to sense failure - and then adapt


Augmented Reality in the Board Room?  Mind Expanding?


One fascinating area of research is Augmented Reality (AR)


Specifically – can Augmented Reality technology provide a net benefit to our executives and board members.   Can it provide incremental value to the meeting in the way a WhiteBoard does - but take it up a notch cognitively? Can it expand mental models?


Augmented Reality and Cognitive Extenders have potential for executive decision support - and can potentially:

  • Invoke and project high-fidelity ideas into the room - or across the country
  • Create or illustrate new mental models; or retrieve them; for better communications
  • Help with stakeholder communications and getting alignment of ideas
  • Engage executives to use more of their 100 billion neuron brains - and extend cognitive capabilities

While it is not yet clear if AI+AR is 'net positive' (overcoming technical drag) this medium clearly has the POTENTIAL to be very disruptive.   Watch this space!



Conclusion - EPS, ROI and JARVIS: 


The combined revenue of US Fortune 500 companies exceeds $12 trillion with combined profits of $840 billion.

A well informed leadership team – with the means to surface the right knowledge at the right times, is going to make better decisions.

Better decisions will result in increased competitiveness and profitability.

The only question is – are we 5 years or 25 years from this being a reality? 

I believe the future is closer than most people think!



Comments and ideas are my own and do not necessarily reflect my employer’s views. 



Interested in more content by this author?

About the Author

Ryan Anderson

Ryan Anderson

Hi! I like to play with data, analytics and hack around with robots and gadgets in my garage. Lately I've been learning about machine learning.

About this blog

Description is...<br/>Data Analytics & Visualization Blog - Generating insights from Data since 2013

Created: July 25, 2014


Up Next