凯捷咨询《解决问题的方法与假设，2004年版，凯捷大学原创》（国
- 1. Structured Problem Solving & Hypothesis Generation
- 2. Goals of this moduleLay out a systematic approach to solving business problems –“Structured Problem Solving ” Establish a common “modus operandus” for Consulting teams Practice the suggested process on a real-life exampleCapgemini - All rights reserved
- 3. Strategy is about thriving in a changing world…“The picture’s pretty bleak, gentlemen . . . The world’s climates are changing, the mammals are taking over, and we all have a brain about the size of a walnut.”Capgemini - All rights reserved
- 4. … which is all about “decision making” “Strategy is about making decisions” The best strategy “makers” are able to blend analytic techniques with an understanding of the future uncertainties and simple good luck Based on often imperfect information they make decisions and then drive implementationCapgemini - All rights reserved
- 5. There are two basic approaches to problem solving; but both can workThe “there's a pony in here somewhere” approachThe structured analytic approachPotential for rich powerful solutionsScurry around analyzing tons of data to see if you can find something usefulGet the dataPotential for good (and mixed) solutionsDefine problem and hypothesesCapgemini - All rights reserved
- 6. Defining the issue is the first step in the journey to final recommendationsDevelop Conclusions and Make Recommendations to ImplementFind InsightsAnalyse DataGather DataForm HypothesesSo what? —aha, new thoughtWhat you should do… and howDefine the IssuesWhat are the questions keeping you awake at night?Factual information gathered to prove or disprove hypothesesAnalyse what the data tells usStatements that provide direction and structure for the analysisCapgemini - All rights reserved
- 7. Hypothesis formation ensures that our analysis is focused on our client’s problemForm HypothesesDefine the IssuesGather DataAnalyse DataFind InsightsDevelop Conclusions and Make Recommendations to ImplementSo what? —aha, new thoughtWhat you should do… and howWhat are the questions keeping you awake at night?Factual information gathered to prove or disprove hypothesesAnalyse what the data tells usStatements that provide direction and structure for the analysisCapgemini - All rights reserved
- 8. Hypotheses are developed in three stepsWhat are the real strategic issues? What is the impact on the organisation? What are the priorities?We think . . . It looks like . . . The right answer may be . . . The options could be . . . We believe this to be true . . . A series of statements, not yet backed by data Based on initial data search or expert opinionsA number of assertions need to be true for a hypothesis to be valid.Is there a market for white label insurance products? Direct channels are growing Retailers have a strong channel and brand There are existing productsChubb Insurance can access new business by using the brand strength of Tesco resulting Is low cost of acquisition and profitable businessDefine the questionReview and Describe Multiple AssertionsForm the hypothesis123
- 9. A hypothesis should identify not only the issue but also the cause and the impactThink through these three stages as you create a hypothesis to help you plan out how you will test it: What is the issue? What is the underlying opportunity? Where is the advantage? What do you think causes the issue? What are the key drivers of the process? What is the impact of the issue? How can we tell there is an opportunity? Why do we care?“x is an opportunity ...”“…due to...”“…resulting in...”
- 10. The Structured Problem Solving Process covers the life cycle of a consulting engagement Define the Problem Structure the Problem Develop a Hypothesis Execute the Analysis Develop a Recommendation Create the Communication Deliver Communication Follow Up with ClientCovered in current moduleCovered later in the weekCapgemini - All rights reserved
- 11. The Basics - Problem Solving ApproachCapgemini - All rights reserved
- 12. DEFINE THE PROBLEMSTRUCTURE THE ANALYSISFIND THE SOLUTIONOur problem solving approach produces results through answering a simple series of questionsIs there a problem or opportunity? If so where does it lie? Why does it exist? What could we do about it? What should we do about it?Fine, but IWIK H2 do this...Capgemini - All rights reserved
- 13. Source: Barbara Minto, “The Pyramid Principle”.Always ask: “Are they Mutually Exclusive and Comprehensively Exhaustive (MECE)?”Logical pyramids are basic tools for this approach, helping you to define, structure and solve the client’s problem 1. Ideas at any level in the pyramid must always be summaries of the ideas grouped below them 2. Ideas in each grouping must always be the same kind of idea, and they must answer the same question implied by their summary 3. Ideas in each grouping must always be in a logical order:Main AssertionKey LineDeductively or inductively Trace course or time order Divide or structural order (e.g., Sales, Marketing, Manufacturing, etc.) Classify or degree order (e.g., most important, 2nd most important, etc.)Capgemini - All rights reserved
- 14. Logical pyramids increase the effectiveness of problem solving, results delivery and communication…Vertical structure Horizontal structure Key line / narrative…but help ensure thorough analysis……while decreasing complexity……and increasing the power of presentations Logical pyramids have simple rules…The effectiveness of our work depends heavily on how compellingly we can argue that the solution we put forward serves the client bestThe reasoning we have to apply to come to our solution is complex and difficult to summarize for brief client interactionsTo build succinct and compelling presentations of our work, we use logical pyramids as the preferred communication style: Pyramids make information more memorable and meaningful They lead to a clearer definition of the problems we solve They structure our solutions to these problems and make them more compellingSource: Barbara Minto, “The Pyramid Principle”.…which will make your work/life much easier to handleCapgemini - All rights reserved
- 15. Define and Structure the ProblemCapgemini - All rights reserved
- 16. “If you don’t know where you are going, any road will take you there.” -AnonymousWhy problem definition mattersCapgemini - All rights reserved
- 17. In structuring a problem, break it into smaller, easier-to-handle components AND start with the right definitionsUS Car MarketLight TrucksPassenger CarsBig 3Mini VansSport Utility VehiclesFordGMChryslerBut be careful — why does this not work?Capgemini - All rights reserved
- 18. "Mutually Exclusive and Comprehensively Exhaustive"The most important rule for any structure you imposeCapgemini - All rights reserved
- 19. DivideClarifyDiagnostic FrameworkCause Effect FrameworkStructural FrameworkProblem StructureTo help you solve a problem, your structure must: Disaggregate the problem into smaller and easier to solve components Be a “MECE” description of the problem and its possible solutionsTrace CauseThere are three ways to structure a business problemSource: Barbara Minto, “The Pyramid Principle”.Capgemini - All rights reserved
- 20. Head HurtsPhysicalMentalExternalInternalStress, TensionHypochondriaBumped, Bruised HeadAllergiesBad Weather, Sinus Headache, Flu, ColdBrain TumorWater on the BrainExample 1 — disaggregate the problem into a diagnostic solution treeSCooPSource: Barbara Minto, “The Pyramid Principle”.Capgemini - All rights reserved
- 21. Store is within shopping radiusdo not know about the storeknow about the storenever visit the storeenter the storedo not buymake a purchasedo not come backmake repeat purchasesLocationAdvertisingSignage, CIConversionCustomer ValueRoot CauseExample 2 (trace cause) — disaggregate the problem into a cause-effect frameworkHow can TESCO improve its sales productivity (sales/sq.ft.)?Capgemini - All rights reserved
- 22. Account typePack sizeREPURCHASE Target market persuaded to repurchase?DISTRIBUTION Brand made available?TRIAL Target market induced to try?AWARENESS Target market aware?Product rejectionPrice/value rejectionFrequency of useOccasion of useConsumer profilePackDisplayPromotion typePromotion tuningSell-in effectivenessFeedbackIf all lines of inquiry fail to reveal a problem source, go back to consider whether target market and consumer benefit have been accurately defined.Attribute awarenessAdvertising recallAdvertising spending rateMedia mixRegional weightAdvertising communicationTarget marketConsumer benefitRegionSales force coverageSales force directionTrade termsExample 3 — disaggregate the problem into an intrinsic structure Why does Wimpy’s not show the anticipated financial performance?ChannelSource: Barbara Minto, “The Pyramid Principle”.Product spec.Selling priceBrand namePackagingPOSITIONING Brand properly positioned for the market?Capgemini - All rights reserved
- 23. Exercise 1 — Kmart vs. Wal*Mart: define and structure Kmart’s business problemSituation Kmart and Wal*Mart operate similar chains of Full line Discount stores Different pricing strategies: Kmart follows a promotional pricing strategy of weekly sales, offering discounts on selected items, Wal*Mart follows an EDLP strategy (EDLP = Every Day Low Prices); Kmart's regular prices are higher than Wal*Mart’s, its sales prices are lower Wal*Mart has a better price perception than Kmart Kmart has a higher GM than Wal*Mart (23% vs. 21%) Kmart has significantly higher SG&A as percent of sales, which eliminate Kmart's Gross Margin advantage over Wal*Mart Wal*Mart’s scale advantage is not driven by the number of stores, but by its sales per store (better sales per square foot) Kmart has significantly lower sales per square foot sales than Wal*Mart ($170 vs $250). With Wal*Mart’s sales productivity, Kmart would be about as profitable as Wal*Mart Complication Kmart is operating at break-even, and Wal*Mart’s aggressive expansion puts more and more of Kmart's stores into direct competition with Wal*Mart, decreasing their store contribution and Kmart's overall profitability Question ?????Structure Kmart's problem to help its management devise a solution, including identifying the key question that our study must answerCapgemini - All rights reserved
- 24. Exercise 1 — define and structure Kmart’s problem: first step is to logically organize the factsCompanyCompetitorCustomerHigher GM than WalMart (+)Higher SG&A expenses than WalMart (-)Aging storesHi / Lo pricing strategyLower net income than WalMartEroding same-store sales vs. WalMartSales / sq. ft. higher than KmartAggressive expansion into Kmart territoriesEDLP pricing strategyCustomers perceive WalMart prices better than KmartCapgemini - All rights reserved
- 25. Exercise 1 — define and structure Kmart’s problem: second step is to iterate to drive insight creationAlthough Kmart’s Hi/Lo pricing strategy leads to high GM, high SG&A has led to eroding sales productivityHigher GM than WalMart (+)Higher SG&A expenses than WalMart (-)Aging storesHi / Lo pricing strategyLower net income than WalMartEroding same-store sales vs. WalMartSales / sq. ft. higher than KmartAggressive expansion into Kmart territoriesEDLP pricing strategyCustomers perceive WalMart prices better than KmartWalMart is aggressively expanding, with highly productive stores and a different pricing strategy than KMartCustomers perceive WalMart delivers higher value in some areasHow can Kmart improve its sales productivity (sales / sq. ft.)?Capgemini - All rights reserved
- 26. Exercise 4 — summary performance data for a credit card issuer… what’s the problem?Note: All figures in 1000s. Assume no price inflation and that interest rates have remained constantThe 80/20 Rule… 80% of the answer is in 20% of the data Often we miss the goldmine because we are busy trying to value the shack built on top of it This data, taken from a real client (but rebased,) tells the whole sorry story of their strategic problem in one picture Capgemini - All rights reserved
- 27. Exercise 5 — using a “quick and dirty” approach can produce surprisingly accurate results30 million? 300 million? 3 billion? 30 billion? 300 billion?How Many (Retail) Litres of Petrol Are Sold in France Per Year?Data: French Population ~ 60 million. 1 Gallon = 3.8 Litres Capgemini - All rights reserved
- 28. Develop a HypothesisCapgemini - All rights reserved
- 29. Explicitly ties your analysis to your problem definitionWhy hypotheses matterThey keep your effortHelps define the level of accuracy that mattersEnsures you analyze no more than is needed to disprove hypotheses within a reasonable doubtAllows quick check before massive data collection and crunching: “If we confirm our belief in the hypothesis, will we be able to act on it?”Keeps you efficientOn targetAccurateMinimalActionableOn TimeCapgemini - All rights reserved
- 30. Source: Barbara Minto, “The Pyramid Principle”.1. Rule2. Case3. ResultRule Case ResultIf we put the price too high, sales will go down We have put prices too high Therefore, sales will go downCase Result RuleWe have put prices up Sales have gone down Sales have gone down because the price is too highSales have gone down Sales go down when prices are too high Probably we have put prices too high HypothesisResult Rule CaseAbduction is a variation on deductive and inductive reasoning and a powerful tool to develop hypothesesDeductionInductionAbduction1. Case2. Result3. Rule1. Result2. Rule3. CaseCapgemini - All rights reserved
- 31. DEVELOP A HYPOTHESIS What differentiates a good hypothesis from a bad one?On target: Answers the core question on the clients’ mind Accurate: Embraces the entire range of competitive or profit drivers Minimal: “Occam’s razor” Actionable: Can be quantified and tested On timeGood Hypotheses AreCannot be refuted Cannot possibly be quantified Require you to “Boil the Ocean” Are so obvious that nobody can intelligently disagree with it Or look like an abstract model of a process consisting of boxes, arrows and cloudsBad HypothesesCapgemini - All rights reserved
- 32. DEVELOP A HYPOTHESIS How to find a good hypothesisThere are proven processes to generate hypotheses: Talk to lots of people about the core problem Have a brainstorming case team meeting Taguchi method — ask “5 Why’s” Start with general validity checks for your hypotheses: Assume perfect rationality — how ought the business system to behave? Look for analogies (other industries or problems) Imagine that you had perfect information — what would the ideal analysis look like? Try Lateral Thinking: Think about the problem from a different angleCapgemini - All rights reserved
- 33. DEVELOP A HYPOTHESIS What if your hypothesis is proven wrong?You Win! Hypothesis-driven thinking requires frequent reviews of the prevailing and current hypotheses: Do we still think they are right? Are we making progress towards disproving them? Are they at the right level of generality/predictiveness? Having proven a hypothesis wrong means you have made a big step towards answering your question!“A wise man sometimes changes his mind, but a fool never.” -Arabic ProverbCapgemini - All rights reserved
- 34. SummaryTo be developed – thank you!!Capgemini - All rights reserved
相关PPT