Building Analytics/Decision Management Capabilities in your organization

I just finished reading the book: The Deciding Factor: The Power of Analytics to Make Every Decision a Winner written by Larry Rosenberger, John Nash and Ann Graham. All three of them were associated with Fair Issac (FICO) at one point.

The authors seem to have great amount of experience in the the field of building analytics and decision management capabilities for their client. Most of the book has generalizations about how to do this. I found the last chapter of the book very useful, other than that it seemed like a listing the facts session.

The last chapter definitely is a good read for someone at a Manager Analytics or Director Analytics/decision management role.

The first thing I am going to list is the high level steps in building the decision management capabilities in your organization. I am only listing the high level steps, please read through the details in the book to get the complete perspective of the authors. All copyrights to the methodology belong to the authors/fico.

Decision Management Methodology
Stage 1: Set Strategy and identify the business opportunity.

• Identify and prioritize opportunities
  
• Assess the scope of the opportunity
  
• Create a high level plan to address opportunities and scope
  

Stage 2: Identify critical decisions and potential decision yield

• Create decision inventory and business process flow
  
• Identify and design pilot model to address the objectives
  
• Capture your baseline decision yield
  
• Quantify potential improvements to your decision yield
  

Stage 3: Design the business architecture for your decision environment

• Determine the best analytic approach
  
• Design your decision environment
  
• Design your decision platform
  
• Define the decision management roles, responsibilities and decision rights
  

Stage 4: Build the data environment required to inform decisions

• Design data flow: sources and sequences
  
• Assess and address gaps
  
• Map connectivity to third-party data providers
  

Stage 5: Build mathematical models to improve decisions

• Gather data required for modeling
  
• Build your models
  
• Test Model performance
  

Stage 6: Build and modify the operational environment to enable decision execution

• Build your decision management application
  
• Build decision rights: organization structure, skills and compensation
  
• Roll out decision process and rules
  

Stage 7: Continually improve the decision environment

• Operate your new decision environment
  
• Confirm realization of your decision yield
  
• Identify and implement changes to your decision environment
  
• Feed new knowledge back into your decision environment
  
• Identify new decisions to improve
  

These are the questions that everyone in a decision management team should be asking to their colleagues across the organization at one point or the other

1. What are the possibilities?
   
2. What are the important opportunities in alignment with company strategy or problem to be solved?
   
3. What are the most important decisions related to these opportunities?
   
4. How should opportunities be pursued based on priorities?
   
5. Making decisions: exactly how long does it take?
   
6. Making decisions: For which product or product line is this an issue?
   
7. Making decisions: Is the delay/issue with decisions across the product lines?
   
8. What are the important decision points?
   
9. Who is making the decisions?
   
10. How are the decisions made?
    
11. What is the current decision performance?
    
12. Can the financial, functional, and technical decision making be improved?
    
13. What are the improvements possible?
    
14. How well do we know our customers?
    
15. Is our strategic segmentation approach integrated into our operating environment?
    
16. Is our customer segmentation granular enough to enable customer specific treatment?
    
17. Are we using analytic capabilities to enhance effectiveness? are we making decisions using judgmental best practices, standardized rules, predictive models, or real-time optimization?
    
18. Who is making which decisions and who executes which functions? based on what criteria and authority?
    
19. How consistently does this play out across channels, product lines, geographies, and so on?
    
20. How coordinated are our marketing activities across different product lines?
    
21. Are our customers or prospects receiving conflicting messages from our organization?
    
22. What is the impact of introducing new strategies into our operating environment? how much retraining is required? What modifications are required in our underlying systems infrastructure?
    
23. Is our organization adept at assimilating change?
    
24. What is the quality of our organizational communication vehicles?
    
25. Are we making decisions through manual intervention or human review, or though scalable, automated rule systems?
    
26. What is our sense of cost versus quality for this area, and what is the optimal balance?
    
27. Are our customers being lost because of ineffectiveness and slow turnaround times in our organization?
    
28. How much time do we need to return a decision to a customer or prospect?
    
29. What is the potential gain of an incremental increase in system processing time for a decision?
    
30. How strong are the current analytic capabilities and what new analytic capabilities are needed?
    
31. What existing models are outdated and need updating or redevelopment?
    
32. Are we periodically updating business logic based on market learning?
    
33. What changes are required to improve the performance of our business logic?
    
34. Do we have the right skill sets and resources assigned to the most valuable priorities?
    
35. What changes to organizational structure, roles, responsibilities, and requisite skill sets have we identified?
    
36. What training and communication vehicles are required to ensure a successful and ongoing rollout of new initiatives and strategies?
    
37. What new information needs have we identified?
    
38. Do we have the right information available at the point of need?
    
39. What additional input data, internal or external, is required?
    
40. What core business process changes or improvements have we identified?
    
41. How will we manage these to ensure consistency, optimal performance and agility?
    
42. What are the best analytic methods to apply to our decision sets?
    
43. Which specific decision areas must we address?
    
44. What capabilities do we need to create and modify?
    
45. What organizational changes do we need to make in terms of roles, responsibilities and structures?
    
46. Could a forecast of future outcomes or customer behavior help us?
    
47. Is manual review of data part of this decision or business process?
    
48. Is the decision complex enough that modeling the results of these decisions could lead us to make better decisions?
    
49. How do we efficiently and effectively integrate the necessary data into our decision environment?
    
50. Is this data accurate?
    
51. What processes are in place to ensure that it is correct?
    
52. What are the methods for identifying and correcting inaccurate data?
    
53. Are we using the best sources of external data?
    
54. Are only partial customer records available for this decision?
    
55. Are we appropriately collecting, logging, and storing data acquired through our transactions and interactions, both to guide later decisions we make and for reporting purposes?
    
56. At what point in the process should we acquire and pay for the data?
    
57. What should the contractual terms be to support long-term use?
    
58. How do we integrate the necessary data into our decision environment, with sufficient service level agreements to meet our processing requirements?
    
59. How much can we improve our analytic performance?
    
60. Which characteristics have the greatest impact on model outcomes?
    
61. Is it operationally feasible?
    
62. Who owns the ongoing management and maintenance of our decision-making environment?
    
63. Who needs to be trained in using these capabilities?
    
64. Are we realizing the improvements we expected?
    
65. Can we identify areas for additional improvement?
    
66. Are these new decision areas we should address?