© 2008 MindTree Li © 2009 MindTree Limited CONFIDENTIAL: For limited circulation only How Dashboards & Visualizations can drive result-oriented decision making Saurabh Jain December2009
© 2008 MindTree Li© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
How Dashboards & Visualizations can drive result-oriented decision making
Saurabh JainDecember2009
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Agenda
Introduction
Dashboard Best Practices MindTree Approach
Metric Alignment
Visualization Best Practices
Design and Navigation
Insights and Learnings Common Mistakes vs. Best Practices
Approach to Performance Management
MindTree’s Approach
Success Factors
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Dashboard
Slide 3
1.Visual
2.Monitor Performance
3.Drill down to lower level
4.Actionable
5.Diagnostic
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only Slide 4
Definition of a Dashboard
“Dashboards are much more than pretty data presentations. They are powerful, flexible business tools which allow for true information democracy – actively empowering business users at all levels to make better decisions every single day.”Hyperion White Paper
“It is a well-established management principle that you cannot manage what you cannot measure. It is equally true, however, that you cannot manage well what you cannot monitor. That is where enterprise dashboards come in.”Business Case for Enterprise Dashboards, Dashboard Execution
Visual elements of performance management system both a process and a
tool
Few sets of indicators
represent central performance areas
Focus on trends and patterns to
monitor performance
Dashboard
Diagnostic means of ascertaining what is working well and what
needs attention
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Tactical/Interactive Dashboards
Operational/Static Dashboards
Operational, detailed data,
process centric operational
reports
Summarized, dimensional data with hierarchies, slice and
dice
Graphical, abstracted data
with graphs, charts
Application of Dashboards ( the three 3’s)
Slide 5
Management
Analysis
Monitoring
Information
Data
Strategy
Implementation
Sources: Wayne Eckerson 2005
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only Slide 6
How Dashboards Help
Static/ Operational Dashboard
Interactive/Tactical Dashboard
Strategic Dashboard
Monitoring Dashboard BI Portal + Dashboard Scorecard
Analysis Standard/Canned Reports
OLAP AnalysisInteractive reportingAdvanced VisualizationScenario Modeling
Statistical ModelsPlanning Models
Management Typically NotAvailable
AlertsWorkflow (Possibly)
Strategy MapsPlanning Models
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
ROI on Dashboards
Slide 7
Return On
Investment
Cost Savings
Business Benefits
• Reduction/Elimination of effort for consolidation of disparate reports
• Reduction in time for coordinating and monitoring complex processes
• Reduction in time wasted in reviewing and assimilating overwhelming amounts of data and reports
• Elimination of redundancies within organization in processing similar data
• Reduction in effort for communicating progress and enforcing compliance
• Better decision making with more current and live information
• Better business insight due to improved data visibility through enhanced data visualization
• Proactive and timely decision making through alerts and exception management
• Greater democratization of information empowering frontline organization
• Improved interactions with external stakeholders (customers, vendors) with transparent communication
© 2008 MindTree Li© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only © 2009 MindTree Limited
How to Build a DashboardBest Practices
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
MindTree’s Approach
MindTree’s Performance Management
Approach
Metric Selection
and benchmarki
ng with Industry
Metrics Balancing
(Right blend of Lead and
Lag indicators)
Metrics Association
“Right” Visual
Construct Selection
Context Mapping
with Metadata &
Analysis notes
Focus on metrics action-ability
(exceptions & alerts)
Supply Related
Analytical context
Dashboard design
validation by usability specialist
Guided Analytics (3
Analysis Levels)
Visualization
Design & Navigation
Alignment
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Alignment - Right Metric
Slide 10
1. Monitoring is important to directly controlling behavior and indirectly to performance - the speedometer reading impacts how hard or soft Tom pushes on the accelerator
2. A few key measurements will go a long way toward keeping a company on track towards achieving its improvement objectives - like those on a speedometer and fuel gauge.
3. Seemingly relevant, but cumbersome, measurements are of little use, and are possibly a hindrance, in helping to improve performance - like the Temp Gauge in the car.
4. Picking the wrong measures and leaving out important ones could lead to situation like losing market share or having poor financial performance is - the way getting speeding tickets and running out of fuel is an expensive way to drive a car.
Key Learning from the story
Common Interpretati
onCredible
Data
Transparent, Simple
Calculation
Actionable
Perfect Metric
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Alignment – Right Balance of Metric
Slide 11
Cost
Valu e
• Leading Indicators• Predictive• Actionable
• Lagging Indicators• Corporate Metrics• Reactive
• Lower-Level Metrics• Small Audience
• Data Integrity Issues• Subjective• New eMetrics
• Financial Metrics Will Typically Be Quicker Hits
• Ops & Quality Measures Typically Take More Time & Money
• Predictive & Market Measures Can Be Game Changers
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only Slide 12
Alignment – Metric Association and Grouping
Type Description User
Enterprise Holistic view of the enterprise from consolidated data across divisions and business segments
Senior Management
Divisional Performance metrics for the division (E.g. Sales, HR, Finance, Purchasing)
Divisional Heads, Operational Managers
Process Monitor business processes or widespread activities
Across multiple user profiles
Application Specific metrics defined for transactional applications within which dashboard is embedded (E.g. Asset management)
Users of the application
Type Description User
Vendor Metrics relevant to vendor Vendor
Customer Outward facing dashboard with metrics relevant to customer
Customer
Employee Metrics relevant to employees Employee
Fun
ctio
nal D
ashb
oard
Sub
ject
/Rol
e-B
ased
D
ashb
oard
Reference: Shadan Malik - Enterprise Dashboards,
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Visualization – Why it is Important
Slide 13
1984 111989 301994 1302000 3002001 4002002 600
2003 11002004 23002005 39002006 60002007 78002008 9500
A Simple Example:
Your organization has grown leaps and bounds.
….The New York skyscrapers on the other hand would register the point very well.
1984 1989 1994 2000 2001 2002 2003 2004 2005 2006 2007 20080
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Year
# o
f E
mp
loyees
Obviously, the table above has little impact…..
Representing data with right visual construct meets the objectives
1. Makes it easy to comprehend data2. Amplifies the impact
You can use the below table to prove the point……
Reference: Vivek Singh – 14 tips to present awesome charts
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Visualization - Which component to choose?
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Visualization - Choosing the right construct
Slide 15
A bar graph showing trend on sales and profitability…can it be better?
Yes! When it is more than 15 data points and a trend, a line graph is more intuitive.
A bar graph showing shareholding pattern of a company
A pie chart would have been a better choice
A pie chart however is not a good choice when comparing the shareholding across multiple companies
Co 1
Co 2
Co 3
A stacked chart would have served the purpose better
Co 1 Co 2 Co 3
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Visualization – MindTree’s Best Practices Framework
Selecting the right visual construct makes analysis faster, clearer and more insightful
Determine Enterprise Dashboard Information needs/KPIs
Segregate dashboard information needs/KPIs by business processes and user groups
Align KPIs with industry best practices
Derive Analysis Scenarios (drill down paths, related reports, events, actions)
Select the right visual construct for delivering insight
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Design and Navigation - Essentials
Slide 17
Ensure compactness modularity, gradual reveal, guide attention, support casual use, lead to action & be customizable
Incorporate drill downs, filters, comparisons, alerts and the option to export or print
Structure can serve as a navigational mechanism for the user. It shows where to start, and where to go next.
Timeliness, Aesthetic value, Mobility, Connectivity, Data detail, Data density, Interactivity, Collaboration
Form Structure
DesignFunctionality
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Design and Navigation – A good example
Slide 18
Functionality- Advance Features
Design Principle
Structure
Form
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Design and Navigation - Key Elements
Slide 19
Standardized Home Pages
Quick Location Navigation
Metrics Navigation
Crisp LineofSight
One ClickDrill Downs
Drill Downs
Automatic“Alerts”
E-MailNotification
ASIA
Dashboard
© 2008 MindTree Li© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only © 2009 MindTree Limited
Insights and Learnings
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Common Mistakes vs. Best Practices
Slide 21
Mistake Best Practice
Pushing too many reports Templates refreshed on demand
Focus on 1st generation data Focus on high value data
Hard-code paths to shallow answers
Free-form Drills, Sorts and Queries to Deep Data
Focus on pretty pictures Focus on Visuals That Quickly Tell An Important Story
Limiting the data model Wide, Deep and Fully-Integrated Data Models
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only
Approach to Performance Management
Slide 22
Identify the right measurement / management methodology (e.g. Balanced Scorecard)
Identify and Implement the right Performance Management Platform
Build Enterprise Data Warehouse
Carry out a successful change project to align users to the new EPM framework
Step 1
Step 2
Step 3
Step 4
Manage organizational change for new management initiative building a performance accountable organization
Train users empowering them with the tools to help them manage their execution towards the companies goals and objectives
Leverage on consolidated, cleansed and validated data from EDW to build a performance management system
Implement the management framework, KPI lineage, strategy maps and plans identified at enterprise and function level
Build scorecards and management cockpits enabling business modeling, what-if analysis etc.
Resolve the integration challenge by providing a central consolidated repository
Automate critical functional & cross functional report
Provide analytical reports for information driven insight
Technology Intensive
Change Management Intensive
Identify clearly the goals and objectives of the enterprise
Identify Management framework best suited for the organization
Identify and formulate KPI lineage, metric definition and strategy maps in alignment with the management methodology and the top line Vision and Strategy
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only Page 23
Therefore … MindTree’s 2-Staged Approach
Stage 2: Bolt EPM Platform on EDW
Stage 1: Build Enterprise Data Warehouse
Transactional Systems
StrategyFramework
(EVA / BSC / ABC)
“Plumbing” IntensiveData IntegrationData qualityPerformanceScalabilityLarge volume handling
“Consulting” IntensiveMetrics RationalizationBest practice metricsAlignment of rolesTying metrics to actionChange Management
© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only Page 24
13 Key Success Factors
Executive sponsorship (Budget)
Give the project a brand name
Deploy incremental projects with a strategic approach – No Big Bang
Start in the highest value areas of the organization
Alignment of metrics to the top line strategy & consistency in metric definition
5-7 indicators on a dashboard is optimal
Should be part of a comprehensive BI strategy
Adopt a single integrated solution with shared, central database
Attractive presentation with meaningful content
Start simple and leverage existing investments
Drive traffic to the solution (HR Benefits)
Over-estimate your data integration needs
Leverage the solution to provide external access to your information
© 2008 MindTree Li© 2009 MindTree LimitedCONFIDENTIAL: For limited circulation only © 2009 MindTree Limited
Q & A