Structuring Systems for Maximum Performance

Presentation regarding system design, originally presented at the 2007 SOA Health Meeting.

Transcript

1. Performance Management according to Gartner Combination of management methodologies, metrics, and IT (applications, tools and infrastructure) that enables users to define, monitor, and optimize results and outcomes to personal or departmental objectives while enabling alignment with strategic objectives across multiple organizational levels (personal, process, group, departmental, corporate or business ecosystem” •Gartner - October 2006 •“Understand Performance Management to Better Manage Your Business” Structuring Systems for Maximum Performance
2. Structuring Systems for Maximum Performance Kevin Pledge, President and CEO, Insight Decision Solutions Carl Desrochers, 2nd Vice President and Actuary, Berkshire Life
3. Kevin Pledge  President and CEO – Insight Decision Solutions  Developer of Business Intelligence system designed specifically for insurance, includes experience studies, EBS and sales analytics.  Consulting and implementation services.
4. Carl Desrochers  2nd Vice President and Actuary, Berkshire Life  Experience with IDI, LTC, Individual Health Insurance and reinsurance.  Currently responsible for major project to overhaul analytics and performance management reporting.
5.  Where are we today?  The Technology Stack  The Business Model  The Analytics Agenda
6. Executives care about Performance Management Accelerate the decision-making process Flexibility to handle changing business conditions Aligned, accountable, and secure PM They can’t achieve this because:  They need to consolidate reports from multiple areas  Conflicting information *
7. Actuaries want to contribute to Performance Management Need a user-friendly and flexible environment Collaborate as part of the performance management process Define, modify, and maintain business rules without IT They can’t accomplish this because:  They can spend 75-85% of their time just managing data  Lack of confidence in data *
8. IT Managers want to enable Performance Management Reduce the complexity, maintenance, and infrastructure cost Empower the business to improve performance Provide access to information while supporting compliance They can’t achieve this because:  They have multiple systems to manage  Local “solution” conflict with corporate objectives *
9. Performance Management Today Yes, online system linked to DW (9%) Yes, online stand alone system (37%) Yes, paper based (18%) No (36%)
10. Performance Management Survey
11. Technology ERP DW OLAP UI POS PAS Claims V’ln CRM Data Store / Data Bus
12. Technology ERP DW OLAP UI POS PAS Claims V’ln CRM UI UI UI
13. Delivery Collaboration:  KPI’s • Measure • Goal • Indicator rules  Analytics  Reporting
14. Technical Issues  Star schemas, snow flaking  Changing dimensions  Data Mastery  Dashboard vs. Scorecard
15. Technology Planning  Plan on success (be careful what you wish for)  Be a good software citizen  Best of breed components or best of breed system?  Don’t dumb down KPI’s
16. Performance Management Today Executives Information Workers Is any of this still relevant? Can I trust the data? Many business models Manual consolidation Different view of the reality Ad-hoc spreadsheets *
17. How PM Should Work Executives Information I can trust and act on Single business model – Integrate processes Business Model *
18. What is a Business Model?  Business Rules and Relationships • Smoker status • Multi-fund  Reflected in multiple systems, but weakest link reflected in all  Must cover scope of Performance Analytics  It is also complex • Needs to cover full scope of business - Multiple systems • Silos need to be broken • Data means different things to different people • Frequency / Timeliness *
19. Policy Lifecycle Agent Lifecycle Original Agent New Agent Pending Active Terminated Business Model – 3 Entities Also  Reinsurers  U/W’ers  Clients  Etc... On Claim Healthy Who cares about changes in agent hierarchy? How is decline rate measured? Pending Claim On Claim Claim Terminated Claim Process Not Paying Premium Paying
20. Typical Issues  Silo Management • Inconsistent definitions • Inconsistent structure • Premature aggregation  Multiple data entry in different systems • Reconciliation issues *
21. Myth: Unifying Definition  Insurance products are not sku’s: • Sales only care about current – in detail • Actuaries need history • Finance have different view • Operations - admin system (plan code)  Not practical to come up with single definition – timing, existing infrastructure (GL, admin)  Need to link, not redefine *
22. Analytics  Redesign based on the tools available  Actuaries used commutation functions in the ’50s (…and are still part of actuarial mathematics!)  Sometimes worth going back to square 1 and re-invent the wheel! *
23. Performance Management Metrics  Profit is the bottom line measure of performance  Earnings by Source (from an enterprise perspective) • Defines key metrics • Other metrics can be linked to EBS • Tied to actual earnings  Must be adapted to your business model • pmpm *
24.  Expresses earnings based on fundamental business drivers: • Investment Income / Expenses • Deaths / Surrenders • Incidence / Recovery • New Business  Originally developed as tool to understand valuation results  Extended to management tool (Collins/Keller 1993) Source of Earnings - Recap
25. Source of Earnings  Enterprise-wide • Applicable to all lines of business  Sum of earnings sources = earnings • Quantify corporate and “policy” income differences • Requires seriatim policy admin, valuation and projection data, together with GL  Slice by policy attributes, time and duration • Monthly profit arising • New business identified by selection not as a source • Relate to, and extend with, supporting analytics
26. Traditional and Management EBS Traditional  New Business Strain  Premium Profit (load)  Interest Profit  Expense Profit  Mortality Profit  Withdrawal Profit  Morbidity Profit Management  Expected Profit (incl exp’ed new business)  New Business Profit  Inforce Profit • Premium Profit • Interest Profit • Expense Profit • Mortality Profit • Withdrawal Profit • Incidence / Recovery Profit
27. Formula Examples  Premium  Investment  Expense  Incidence  Recovery    Start v Cashflow PP PP    Start v Start i R i R 0 0    Start Start v UE UE                 Healthy Start v v Disabled Entry Healthy Exit R R d R R 1 0 1 0                 Disabled Start v v Healthy Entry Disabled Exit R R r R R 1 0 1 0
28. Consolidated Report Format Source Experience Margin Total New Business X X X Inforce Business Premium X X X Interest X X X Expenses X X X Incidence X X X Recovery X X X Withdrawal X X X Total Actl – Expd Earnings Expected Earnings Actual Earnings Expected Profit on Inforce Business Impact of New Business Experience Gains and Losses
29. Earnings by Source  New business impact • Relate to sales volumes / target – agent production / recruitment • New business distribution  Profit by Sales office / Business Line • Focus on components they influence  Pricing • Actual Profit Signature
30. Experience Studies  Actuarial studies vs. PM metrics • Serve different purposes • Immediate knowledge of results • Results by agent / office etc  Exits / Ave Inforce probably sufficient • Similar to central method • Easier to explain, but timing difference  Agent Persistency Measures • Cohort approach – current ratio faux pas *
31. Claims Analytics  Incurred / Paid • Incurred includes the estimates of future costs • In early durations, much of results based on actuarial results • Source of arguments with Field Force (and sometimes DOI) *
32. New Business Analytics  Throughput  Decline rate Vs. “offer” *
33. Projection Models  Delays in meaningful metrics *
34. People  NextGen Software or Software for the NextGen?  Need vision to look at big picture  Someone that can stand up to the current establishment (a.k.a. silo management) NextGen LastGen Raised on technology Suspicious of it Not afraid of software, but very critical Resist software solutions, selective Know how to sew things together Look for all-in-one solutions Seek a technology solution first Seek a “go-to-guy” first
35. The Future  Convergence  No Programming  Mobile
36. Performance Management Isn’t:  Stand–alone KPIs, dashboard, scorecard  A software solution Is:  Collaborative process  Supported by business model
37. Thank You  Kevin Pledge, kpledge@insightdecision.com 905-475-3282, www.insightdecision.com  Carl Desrochers carl_desrochers@berkshirelife.com
38. Analysis  the abstract separation of a whole into its constituent parts in order to study the parts and their relations [ant: synthesis] Synthesis  the combination of ideas into a complex whole [ant: analysis] Source: WordNet®3.0
39. Volatility of Results  Can’t avoid, issue how to resolve? Still indicate trends. Manage and understand – time trending e.g. recession  One number doesn’t tell the story – need to look at trending. People want numbers e.g. Profit by sales office  Rolling average, trend within is important – disincentive if weighing is applied to stabilize metric