Given at Realcomm, 2009, this presentation covers:
* Technical detail behind a business intelligence implementation
* Building the business case to support a comprehensive business intelligence program
*Using data mining and predictive analysis to understand potential future portfolio trends
2. Agenda
• Overview
• Data Integration, Data Warehouse and Data Marts
• Reporting and Analytics
• Building the BI Business Case
• Possibilities for Data Mining and Predictive Analytics in Commercial
Real Estate Portfolios
Slide 2
3. Overview
• Exploring the technical detail behind a BI implementation
• Building the business case to support a comprehensive business
intelligence program
• Using data mining and predictive analysis to understand potential
future portfolio trends
Slide 3
4. What Should a BI Solution Provide?
• Data transparency, allowing drill through from summarized information
down to the underlying detail
• A platform for monitoring and enforcing data quality standards
• Resiliency to underlying system change
– As underlying transactional systems change the users of the BI
platform are shielded from that change
• Graphical representation of analytics providing immediate
understanding of business trends
• A platform for orchestrating the movement of information between
systems
• A time sensitive view of information across systems
Slide 4
6. Integration, Warehousing and Data Marts
• Data Integration/Warehousing solutions are comprised of:
– Data Dictionary
– Logical Data Model
– Physical Data Model
– Data Quality
– Data Synchronization
– Data Movement Capabilities
• Make sure this is implemented along with a data governance
mechanism and an ongoing monitoring program that ensures
consistent data quality
Slide 6
8. Taking a system agnostic approach to a data model
OSCRE Hybrid Approach
Slide 8
9. Reporting and Analytics
• Reporting in the complex world of commercial real estate can be characterized
as follows:
– Most companies use several dozen Excel spreadsheets to analyze and
report data
– Data is typically scattered in multiple and disparate sources
– “Plain vanilla” reports such as Balance Sheets and Income Statements
are relatively easy to produce at an aggregate level but more detailed
reporting can take weeks to pull together
• The solution
– Find an implementer and vendor who can be relied on to give you what
you really need based on true business requirements
– Consider standardizing on a single technology stack
– Make sure your internal resources understand what the vendor is doing
Slide 9
10. Building the BI Business Case
Building the BI Business Case:
• Quantify Cost Savings
– Interview business users to understand the time it takes to produce the
current reporting and analytics within your organization
– Apply an internal hourly rate
• Quantify BI implementation and ongoing costs
– Consulting costs, infrastructure costs, internal costs
– Training costs
• Determine ROI/Payback
• Simple, right?
Slide 10
11. Building the BI Business Case – Not so fast
• Simple ROI business cases only work in environments where there is a
general consensus that BI is an essential part of the overall organizational
architecture
– Understanding that a transactional system is not a good basis for a data
warehouse
– A system agnostic data and reporting platform is critical to maintaining
business operations
– A potential for expanding to additional asset classes to get a true picture
of an overall investment portfolio
• The qualitative components behind the BI Business Case are
unfortunately the most compelling for implementing an end-to-end
infrastructure
Slide 11
12. Business Benefit of BI
•
Lowers operating costs as a result of eliminating manual
process
•
Reduces the chance of reporting errors
•
Improves the speed and efficiency at which a company can
determine specific exposure and risk, improving overall
business agility
•
Streamlines operations by automating and standardizing the
aggregation of information from various entities irrespective of
geography, technology or business model
•
Establishes an architecture that will support future growth
including additional assets in existing entities, new products and
new platforms
Slide 12
13. Data Mining and Predictive Analytics for Commercial Real Estate
• Used forever by insurance companies to build risk and premium
models
• Takes historical information to predict future trends
• Requires a robust data environment (multidimensional) to be able to
support the analysis
• Technical resources must be able to determine the application
algorithm to apply to a data set
• Results must be aligned to significant macro indicators – examples:
– Economic environment (inflation, employment, rate of economic
growth)
– Regulatory environment
Slide 13