Büyük Veride Dikkate Alınması Gereken 4 Sorun | Big Data: Four Problems to Consider

© 2017 Milo Jones

Big Data: Four Problems to Consider

Prof. Milo Jones IE Business School

© 2017 Milo Jones

Who am I?

» I am:

American; grew up in New York, Wisconsin & Virginia

BA Northwestern, Art History

US Marines; then archaeology, then broker at Morgan Stanley in NY

London Business School MBA

Accenture, London: Organizational Strategy and Change

Dot.com startup; Masters & PhD in International Relations

PhD focused on intelligence analysis, strategic surprise and internal culture of the CIA

Now: I live in Warsaw; I run a consulting company called Inveniam Strategy; I am a non-Executive Director of US FMCG company; I teach at IE about eight weeks a year.

© 2017 Milo Jones

Who am I?

© 2017 Milo Jones

Big Data = Wisdom?

© 2017 Milo Jones

The Role of Theory

© 2017 Milo Jones

Big Data = Wisdom?

» A story from 2012…

© 2017 Milo Jones

What do we mean by “Big Data”?

» Not just everyone, but everything is generating (or will be generating) recordable data

» We may be at another moment when we can “see the invisible”

© 2017 Milo Jones

Problem One

We have the wrong definition of “smart”

© 2017 Milo Jones

Problem One

» We have the wrong definition of “smart”

» Students are taught to give good answers, but not to ask good questions!

© 2017 Milo Jones

Problem Two

“Collection” of data does not prevent strategic surprises

© 2017 Milo Jones

A little reminder

© 2017 Milo Jones

The Lessons of Pearl Harbor

© 2017 Milo Jones

Is collection of data the problem?

» Hypotheses are used to sort “Signal” from “Noise”

© 2017 Milo Jones

The “normal” intelligence cycle

Tasking

Collection

Analysis

Production

Dissemination

The Intelligence Cycle

This is almost never the problem!

© 2017 Milo Jones

W. Edwards Deming, revisited

» Deming: “Without data you’re just a person with an opinion.”

» When Big Data is everywhere: “Without an opinion you are just another person with data”!

» You need hypotheses (i.e. questions) to sort “signals” from “noise”

» Surprises come from “Failures of imagination”

© 2017 Milo Jones

“Failures of imagination” – no one to blame?

© 2017 Milo Jones

Problem Three

Hypotheses and “Failures of Imagination” do not come from nowhere

© 2017 Milo Jones

But does that explain enough?

» What can we learn from “Cassandras”?

© 2017 Milo Jones

The identity and culture of the CIA

© 2017 Milo Jones

Tasking

Collection

Analysis

Production

Dissemination

A New Intelligence Cycle

Hypotheses

Your Identity and culture shape these

Identity and Culture shape Hypotheses (and Surprises)

© 2017 Milo Jones

A sad example

© 2017 Milo Jones

Who is a Cave Man?

» Implication: diversity is a practical concern!

© 2017 Milo Jones

Problem Four

“Centaurs” are spreading

© 2017 Milo Jones

1997

©2015 Inveniam Strategy

© 2017 Milo Jones

2011

©2015 Inveniam Strategy

© 2017 Milo Jones

2016

© 2017 Milo Jones

Problem 4: The Rise of Centaurs

© 2017 Milo Jones

Problem 4: The Rise of Centaurs

© 2017 Milo Jones

» “A weak human + a machine + a good process is superior to…

» To a strong computer alone and…

» (more remarkably)…

» To a strong human + a machine + an inferior process.”

» Big Data + AI = Centaurs in every field of business

» But the challenge is more complex than “analytics”

Garry Kasparov “The Chess Master and the Computer”

Problem 4: The Rise of Centaurs

© 2017 Milo Jones

Practical Implications

© 2017 Milo Jones

Practical Implications

» Job One: remove friction in the interface between your team and their machine/Big Data assistants!

» Job Two: make sure you design processes to create data sets worthy of intelligent machines: polluted or suspect data wastes everyone’s time and people debate data, not strategy!

© 2017 Milo Jones

» Attack exceptions/ interrogate anomalies

» Get good at structuring rigorous discussion of soft data

» Constantly refine the soft skills - inspiring your associates, empathizing with customers, developing talent, etc.

» Giving context to small-scale machine-made decisions is likely to be vital to maintain team morale

Practical Implications

© 2017 Milo Jones

Summary

© 2017 Milo Jones

Summary – The Challenge of Big Data

» Problem One: Being smart used to mean giving “good answers”; increasingly, it will mean asking good questions

» Problem Two: Collecting data does not prevent surprises

» Data rarely “speaks for itself”, it needs hypotheses

» Problem Three: Both hypotheses and “Failures of Imagination” come from your culture and identity

» That means that diversity is a practical issue for identifying both opportunities and risks

© 2017 Milo Jones

Summary – The Challenge of Big Data

» Problem Four: Meanwhile, “Centaurs” (teams of humans and computers), are using Big Data and moving into every field of management

» Management needs to focus the processes of human and machine interaction – this is more than “analytics”

» Once you have clean data…

» Soft skills will be increasingly important for leaders!

© 2017 Milo Jones

Thank you!

Please Connect: [email protected]

Linkedin: https://www.linkedin.com/in/inveniam

Twitter: https://twitter.com/Inveniam