O r g an izat io n al In t ellig en ce A ssessmen t W o r ... · • 3 analytics projects with...
Transcript of O r g an izat io n al In t ellig en ce A ssessmen t W o r ... · • 3 analytics projects with...
JPK
Gro
up
Organizational, Customer & Data Intelligence and Analytics Forum
June 8-9, 2017 • Boston, MA
Organizational Intelligence Assessment Workshop
June 8, 1:00 pm
Founder and Managing Director of Aurora Predictions, a software company that provides revenue foresights planning to ensure companies make their quarterly revenue targets and insights to understand what drives their revenue dynamics.
He is also founder and president of Z Concepts, a company that promotes innovation in technology and civics; and co-founder of US Vigilance
(www.usvigilance.com), inspiring responsible citizenship through fact based and reasoned discussion on public policy. He has been a high-tech entrepreneur for
25 years founding and growing software companies in telecommunication, manufacturing & distribution, high data availability, and predictive analytics.
View presentation online at: https://jpkgroupsummits.com/attendee6/
Presenter: Robert Zwerling – Aurora Predictions
Learn the expected outcomes to compare your organization & assess the viability of the practical use of business analytics
Organizational Intelligence Enablement
Workshop
“Discover your tomorrow today”
Notice
This report contains forward looking statements. Forward looking
statements have no guarantee of being achieved. This
presentation is for information only and Aurora Predictions, LLC
makes no representation or warranty to the accuracy or
completeness of the data herein or fitness or merchantability of
its use of any kind. Trademarks are the property of their
respective company.
/ 2 /
Topic & Goal
Organizational Intelligence (OI) Assessment
Workshop
Benchmark and plan intelligence and data governance
strategy through the systems, people and processes
needed for the enablement of analytics that support OI.
Goal – take home plan to enable Analytics for OI
/ 3 /
About – Robert J Zwerling, P.E.
Managing Director, Aurora Predictions
/ 4 /
• Co-Author, Vigilance The Price of Liberty
• Co-Founder, US Vigilance, promoting responsible
citizenship (www.usvigilance.com)
• Co-Producer with Richard Nixon Foundation, The
Collegiate Forum
• 25+ years growing software companies in M&D, Telecom, High Data
Availability & Predictive Analytics with 3 successful exits
• Authored a dozen papers on predictive analytics, 1 patent pending
• Bachelor & Masters in Engineering, Member Tau Beta Pi, and
Register Professional Engineer
Workshop Plan to “Round the Bases”
• Start with a line drive to right with OI Fundamentals
• Head to 1st base with Building Analytical Thinking
• Round 1st with the Barriers to Analytics
• Steal 2nd with cases of Analytical Failure & Success
• Run to 3rd with Analytics Infrastructure Considerations
• Round 3rd with Analytic System Considerations
• Tag home plate with Building Blocks of DG
/ 5 /
Organizational Intelligence (OI)
Fundamentals
/ 6 /
“line drive to right field”
Four High Level Components of OI
/ 7 /
Culture
What is the Definition of OI
• Organizational Intelligence (OI) is the
capability of an organization to
comprehend and conclude knowledge
relevant to its business purpose
/ 8 /
Understanding the Components of OI
• Comprehend & Conclude – is the process
of analysis, evaluation and conclusion
• Knowledge – is derived from data
transformed to meaningful information
• Data – is relevant (right) data vs. big data
/ 9 /
To Comprehend and
Conclude from Knowledge
of the Data requires
Analytics
/ 10 /
Definition of “Analytics”
/ 11 /
• Reporting: presentation of past data & its
arithmetic comparison – Scorecards/Dashboards
of past are NOT analytics
• Analytics: arithmetic, statistic and mathematic
calculation on past data for insight and prediction
Meaning of “Insight” & “Foresight”
• Insight – revealing facts of the past –
interesting but not necessarily actionable
• Predictions (Foresight) – calculation on
facts of past for future outcomes both
interesting & actionable
(forecasts, predictions, probability & accuracy)
/ 12 /
Data Requirements
Timely
Accurate
Complete
Accessible
/ 13 /
Plan for Implementing Analytics for OI
• Develop Analytical Thinking
• Identify Barriers to Analytics
• Build Infrastructure for Analytics
• Define DG Framework for Analytics
/ 14 /
Building Analytical Thinking for OI
/ 15 /
“heading to 1st base”
Analytical Thinking
• You don’t need to be a Data Scientist
• Analytical thinking is a process approach
to the gathering the pertinent facts and
evaluating those facts
• The goal of analytical thinking is to solve
issues that optimize decisions
/ 16 /
Process of Analytical Thinking
/ 17 /
Task Targets
What Is being solved
Why Is solving valuable
Where Do the pertinent facts reside
Who Owns the facts
When Are the facts updated
How Will solving be implemented & measured
Building Analytical Thinking
• Conclusions and recommendations follow
analysis and evaluation of facts
• Beware Excel for Analytical Thinking
o Conclusions & recommendations lead
analysis and evaluation of facts
(justifying decisions already made)
/ 18 /
Barriers to Analytics & Components for
Success
/ 19 /
“rounding 1st base”
General Perception of Barriers to Implementing Analytics
• Complex
Simpler new generation tools
• Cost not Budgeted
ROI on analytics is 10X
• CEO Support
Management yes, CEO no
• Data Availability/Integrity
Only matter of effort & new gen tools address
/ 20 /
The Biggest
Barrier to Analytics
Is
Culture
/ 21 /
What is the Meaning of “Culture”
• The Merriam-Webster dictionary defines Culture:
Beliefs, customs, arts, etc., of a particular society,
group, place, or time
• The culture needed for “Analytics” is:
Mindset, system, people and processes in the
business to make data driven decisions using
quantitative analysis and predictions
/ 22 /
Components of Analytics Culture & Relation
/ 23 /
mindset
system
people
process
Mindset vs. Analytic Contribution to Business
Reporter
Commentator
Advisor
Partner
Support
Insight
Influence
Impact
/ 24 /
Analytics Contribution
Mindset to Business
mindset
system
people
process
Mindset vs. Deliverables & System
Mindset Deliverables System
ReporterTimely & Accurate
Historical ReportingExcel & BI
CommentatorReporting with Simple
Analytics of Interest
Discovery &
Visualization (DV)
AdvisorInfluential Reporting with
Insights
Desktop Stat (DS) or
DV & Analytics (DV&A)
PartnerImpactful Insights &
ForesightDV&A
/ 25 /
mindset
system
people
process
Types of Analytics Systems
/ 26 /
System Description
Desktop Statistical
(DS)
Excel add-on experimental & personal use,
small database, need statistician
Enterprise Data
Mining
Very complex, large, for specialized use
only, all else suicide
Discovery &
Visualization (DV)
Deep dive with trends & dashboards for
insights but not predictive
Discovery,
Visualization &
Analytics (DV&A)
Deep dive trends & dashboards for
insights & predictive analytics for foresight
mindset
system
people
process
Analytics Implementation Model
/ 27 /
Reporter
Commentator
Advisor
Partner
Support
Insight
Influence
Impact
Analytics Contribution
Mindset to Business
Excel
DV
DS or DV&A
DV&A
System
Analytics
Reporting
mindset
system
people
process
The People & Processes
• People: need threshold number of “core” &
“consumer” users or be at risk to attrition
• Processes: written procedures for regular
operation for data loading to collaboration
Data Accessible
Data Timely
Data Accurate
/ 28 /
mindset
system
people
process
Analytics Implementation Model
/ 29 /
Reporter
Commentator
Advisor
Partner
Support
Insight
Influence
Impact
Analytics Contribution
Mindset to Business
Excel
D&V
DS or DV&A
DV&A
System
Analytics
Reporting
Sufficient People & ProcessesInsufficient
mindset
system
people
process
Case Examples of Analytics
Failures & Success in:
Telecommunication, Healthcare, Retail &
Transportation
/ 30 /
“stealing 2nd base”
Examples of Cultural Deficiencies –
Death by Excel Culture
Fortune 500 telecom company
management believe forecasting &
analytics would enable better
productivity and decisions but failed
against culture of spreadsheets
“people who care don’t know & people who know don’t care”
/ 31 /
Examples of Cultural Deficiencies –
Death by Excel Culture
• 3 analytics projects with proven 10X-70X ROI
o Forecast demand accuracy increased to 98.6%
vs. 94.1% over 1 year or 35 million units delta
o Forecast EDA use 72% of time more accurate
then internal & 71% with 90%-99% accuracy
o Forecast disk utilization and disk allocation to
save $7 million per month
/ 32 /
Examples of Cultural Deficiencies –
Death by Excel Culture
Projects terminated because Excel:
o Protects my job from productivity
o Hides mistakes – just blame the data
o Makes me indispensable – I’m the MESS
/ 33 /
Implementation Failure – Mindset &
System Mismatch
/ 34 /
Management
Staff
mindset
system
people
process
Examples of Cultural Deficiencies –
Death by Reporting Mindset
$5B healthcare provider wants
analytics for better decisions &
personnel development but failed
to develop personnel
“Good enough is good enough”
/ 35 /
Examples of Cultural Deficiencies –
Death by Reporting Mindset
FP&A successful in regular monthly and ad-hoc reporting but
despite repeated attempts failed in analytics
• Comfortable with copy & paste from DV&A to Excel and
email in “a minute”
• Comfortable with a large team spending 2 weeks in C&P to
produce monthly management reports
• Disconnect as Finance says it makes reports Operations
wants & Ops says it gets reports Finance provides
/ 36 /
Examples of Cultural Deficiencies –
Death by Reporting Mindset
Management failed to:
• Push folk from their comfort zone
• Provide time to develop analytical skills
• Put incentive to develop the analytical skills
• Coordinate Finance and Ops analytical needs
/ 37 /
Implementation Failure –
Mindset & Analytics Mismatch
/ 38 /
Ma
na
ge
me
nt
Staff
mindset
system
people
process
Examples of Cultural Deficiencies –
Death by Attrition
NASDAQ Retailer installed analytic
system for better operational
forecasting & planning but failed
to assure sufficient users
“one & done”
/ 39 /
Examples of Cultural Deficiencies –
Death by Attrition
• Goal to reduce monthly planning cycle effort by
50% and get 98% forecast accuracy – goal met!
• One person in Finance trained as “core” user
then promoted but left no written procedures
• Replacement had no processes or consumers
/ 40 /
Examples of Cultural Deficiencies –
Death by Attrition
• People turn-over and usage delay will
doom analytics unless analytics is:
o Quickly integrated into operations
o Has written procedures
o Has sufficient core users & consumers
/ 41 /
Implementation Failure – Insufficient
People & Processes
/ 42 /
mindset
system
people
process
Example of Cultural that Works –
Doing All the Right Things
NYSE freight transportation
company successfully implemented
forecasting & analytics for better
revenue planning & risk mitigation
“putting all the pieces together”
/ 43 /
Asset Discussion
Mindset
Finance sought to partner with business for
analytic revenue forecasting, quarterly 18
month rolling rev planning to mitigate risk &
to manage street expectations
Analytic
System
Realized Excel & DS are not enterprise so
went DV&A
PeopleStart team of 3 core users that grew to 80
consumers in executive, sales & marketing
ProcessUsers trained, operational processes for
data loading & collaborative input
Cultural that Works – Doing it Right
/ 44 /
mindset
system
people
process
Example of Cultural that Works – Doing it Right
• Increased forecast accuracy & forecast line of
sight through analytics & human collaboration
• Extensively use forecasting & correlations with
economic indicators to assure revenue trends
• Extensively use advanced analytics for Street
earnings management
/ 45 /
Example of Cultural that Works – Doing it Right
Use correlations to
find lead economic
indicators &
advanced
forecasting to reveal
risk points in
demand
Units vs. U.S. GDP
Lead 9 Months
Point of discussion
when forecast bar
outside green line
/ 46 /
Item to Note Regardless of Failure or Success
• Data was never an issue
• Big-Data vs. Right Data
• Real Time vs. Right Time
/ 47 /
Case Summary – Killers to Analytics
/ 48 /
mindset
system
people
process
Infrastructure Considerations for
Implementing Analytics
/ 49 /
“running to 3rd base
Checklist for Analytics Infrastructure
/ 50 /
mindset
system
people
process
Analytical Systems – Vendor Sampling
• Desktop Statistical• Crystal Ball (Oracle), Minitab®, Forecast Pro
• Enterprise Data Mining• SAS, SPSS (IBM)
• Discover & Visualization• New Gen Cloud: ClearStory, DataHero, Domo
• Old Legacy: QlickView, Tableau
• Discovery, Visualization & Analytics• New Gen Cloud: Alteryx, BigML, Datameer, Aurora
• Large Cloud: Data Analytics (Oracle)
/ 51 /
mindset
system
people
process
The People for Implementing Analytics
The Evangelist – man with the cause(Man can fly)
The Backer – man with the budget(I’ll invest in that)
The Test Pilot – smart guy(I’ll fly that)
The Congregation – group of users(We’re the airline to fly that)
/ 52 /
mindset
system
people
process
Empirically Deduced – Threshold
of People for an Analytical Culture
You need >3 core users &
>7 consumers, anything
less is at risk from attrition
/ 53 /
mindset
system
people
process
Key Processes – Must be Written
/ 54 /
Processes Goal Benchmark
Data LoadingAssure Right Data
Updated at Right Time
Sufficient
Reaction Time
Data Quality
Assurance
Right Data Without
Errors
Free from Errors &
Missing Data
CollaborationAuthorities for Inputs,
Approvals & Updates
Assure Access &
Advancement
TrainingAssure Capability &
Effectiveness of Use
Clear &
Comprehensive
mindset
system
people
process
Risk of Failure – System, People & Processes Concurrent But
Not Coincident
/ 55 /
People
Processes
System
Small
Area for
Success
Alignment for Success – System, People &
Processes Conncurrent & Coincident
/ 56 /
People
System
Processes
Considerations of
Analytical Systems
/ 57 /
“rounding 3rd base
Ad-Hoc Point vs. Systematic Solutions
• Point solution often require a Data
Scientist(s) to coble together technology
to answer a specific issue
• Systematic solution uncovers heretofore
unknown answers and inspire heretofore
unasked questions
/ 58 /
Systematic & Transportable Solution
1. Systematic – the application of predefined
comparisons and statistics meaningful to
business applied to all data across all dimensions
2. Transportable – use of predefined comparisons
and statistics across any business database
(vs. point solutions are not systematic nor typically transportable)
/ 59 /
Dimensionality & Visualization Intelligence
• Analytic intelligence require unrestricted
dimensionality to reveal value in the data
(why Excel & BI have limited value)
• Visualization can further reveal value in
data – but separate sizzle from the bacon
/ 60 /
Dashboard Sizzle – Good Looking Insightful But Not Predictive
or Necessarily Actionable
/ 61 /
Will this
good
trend
continue?
Visualization Complexity – Lot’s of Value But Need to be Skilled
User
/ 62 /
Presentation
Of
Analytics
Systemization &
Dimensionality
/ 63 /
Analytics Benchmarks
/ 64 /
Analytics Benchmark – CTV
• Cost, Time & Value (CTV) – is the
quantitative measure of analytics
o Cost – how much
o Time – how long
o Value – the ROI
/ 65 /
Data Governance Building Blocks
/ 66 /
“tagging home plate”
DG Primer for Your Information Only
• RWDG is a
comprehensive,
organized & disciplined
primer for DG
• But it’s large & maybe
too much overhead for
some organizations
/ 67 /
https://www.youtube.com/watch?v=cYuRu6P8ugI
“Real World Data Governance (RWDG)”
Definition of Data Governance
Management process of data to
assure it’s timely, complete, accurate
& accessible for reporting & analytics
to its intended business purpose
(DG vs. Data Security)
/ 68 /
Data Governance Actions
• Management – authority & accountability
• Process – written roles & responsibilities
• Timely, Complete, Accurate & Accessible
– are the measurable data goals
/ 69 /
High Level Framework for Data Governance
1. Cornerstones
2. Organization
3. Processes
4. Benchmarks
/ 70 /
Data Governance Cornerstones
1. Management makes personnel time & incentive
to attend to the actions of DG
2. Management uses the analytics requiring DG
3. Defined roles & responsibilities memorialized in
written procedure
4. Defined DG measurements to meet its goals
/ 71 /
DG Organization
• DG is best organized when distributed
• DG responsibility should be with the
users & creators of the data – not IT
• Responsibilities should be part of a
user’s role & not a role by itself
/ 72 /
IT Organization Ground Rules for DG
• Users responsible for data but IT for security
• IT should not erect “high” barriers to get data
o Person – no armies, 1 interface per database
o Processes – no bureaucracy
o Promptness – no delay for data access
/ 73 /
Process is Imperative – Written – KISS
/ 74 /
1 High Level Process Overview
Sub Process Detail
Quarterly Production Update Schedule
Production Update Process - Detail
2
3
4
• Example of Contents of a DG process
/ 75 /
Load
ModelAudit
Model
Backup Model Copy Model
to Sandbox
Analytic
Load
BI
Archive
Analytic
Load
Validate
Backup Model
Post To
WebWeekly
Process
A B C
Critical Path
1 Load Sync Up
- All team members must
participate to achieve
success
2 BI Archive
- Application update
gated by BI archive
3 Audit Model
- Audit and approval of
data load must be
completed prior to
copy to Web
- Time Sensitive
DSub - Processes
1
Sequential Update Process
2
3
4
5
6
Load Sync Up: Align on plan content
BI Archive: Process gate. Data baseline. Load Sync up 5 days prior
Validation/Back Up: Generate initial models for review
Load Model : Each knowledge site populated for validation pre WebReporter
Audit Model: Complete validation to ensure accuracy of data viewed by users
Post To Web: Users provided with access to updated data
Model Advancements
1 System & Report update requests reviewed monthly to support Model Advancement
High Level Process Overview Example
/ 76 /
Sub Process D – Weekly Detail Example
1 Process Implementation
2 - Follow up with data owner will be critical to ensure data across
systems stays aligned
BI Update
Update Web
BI Update Analytic model & BI alignment notify BI data owner of change required
- Successful repetition required to drive successful
implementation
Cost Analyst MC MC
Cost Analyst MC MC
Cost Analyst MC MC
Cost Analyst MC MC
Update Web
/ 77 /
Quarterly Production Update Schedule Example
MK
MV
MC
MC
BI Cost Lockdown / Archive
Ensure model modifications in place
Create & validate load file for each model based on agreed path
Each owner identify method of attribute review
BI alignment
Notify attribute owner of required changes to BI
Backup audited load file & post to Web
BI Archive
/ 78 /
Production Update Process – Detail Example
Finance Cost
Process
LTF Process
Web Updates
Production Updates
/ 79 /
Finance Cost Process ExampleCadence Quarterly per Cost Calendar: Dec, March, June, Sept
Model Attach Wafer Costing
Method
Load File (CSV)
Strat Marketing ForecastUnit Volume From Attach Model + QCA TM1 Feed
Build Demand from Wafer Model + Selected Cost
Inputs from TM1
Ownership
MC: Prepare and load file
OH/JL: Select TM1 Inputs to include in load file
When
Unit Volume: Update with latest Forecast available at the time - per Strat Marketing forecast calendar
Meet a week before cost archive due date and
determine which data will be transferred from TM1 to ISIS, Load file once Finance
cost models are locked
BI FeedBI Feed
Line Analytics Model Component Analytics Model Costing Model
BI inputs
be transferred from BI
to model load file once
Finance cost is locked
Load File
Data Procedures – Detail Needed
• While DG processes can be small &
simple, data loading needs detail
• The benchmark of a good procedure is it
can be used by anyone with little training
/ 80 /
Data Procedures – Detail Needed
/ 81 /
DG Benchmarks
Goals Deliverable Benchmark
Data
Timeliness
Right Time
Availability
Sufficient Reaction
Time
Complete &
Accuracy
Right Data
Without Errors
Free from Errors &
Missing Data
Data
Accessibility
Consumable by
All Who Need
Ease of Access &
Use
DG ProcessWritten Processes
& Procedures
Comprehensive,
Simple & Usable
w/o Training
/ 82 /
Closing Thought on Analytics for OI
/ 83 /
“Know thy self, know thy
enemy. A thousand
battles, a thousand
victories.” – Sun Tzu
Analytics Culture for Success – Know Yourself
/ 84 /
“and the winners are”
Organizational Intelligence
Workshop
www.aurorapredictions.com
“Discover your tomorrow today”
© 2017 Aurora Predictions, LLC & Robert J Zwerling. No portion of this presentation can be reproduced without this copyright notice. Forward looking statements have
no guarantee of being achieved. This presentation is for information only and Company or Author makes no representation or warranty to the accuracy or completeness
of the data herein or fitness or merchantability of use of any kind. All trademarks are the property of their respective company.